House Small Business Committee Interim Report 2024

Small Business:
Instruments and Casualties of the

Censorship-Industrial Complex
The House Committee on Small Business

Interim Staff Report 2024
1
Table of Contents
I. Executive Summary..................................................................................................... 4
II. Introduction.................................................................................................................. 5
III. Committee Investigation.............................................................................................. 8
IV. Small Business: Instruments & Casualties of the Censorship-Industrial Complex..... 9
1. Global Engagement Center Origin & Mission Expansion..................................... 10
2. The Technology Engagement Team & Private Sector Engagement.......................11
3. Awards................................................................................................................... 13
A. Park Advisors..................................................................................................... 14
i. Testbed Methodology & Disinfo Cloud......................................................... 15
ii. Tech Demo Series.......................................................................................... 20
iii. International Tech Challenges........................................................................ 22
B. Global Disinformation Index............................................................................. 24
i. About the GDI................................................................................................ 24
ii. Products......................................................................................................... 26
a. Reports......................................................................................................... 27
b. Dynamic Exclusion List............................................................................... 29
c. Veracity.ai..................................................................................................... 30
iii. Funding.......................................................................................................... 31
a. GEC Funding................................................................................................ 31
b. NED Funding............................................................................................... 33
iv. Partnerships.................................................................................................... 36
v. GEC-GDI-Blacklist.docx............................................................................... 37
C. NewsGuard........................................................................................................ 42
i. About NewsGuard......................................................................................... 42
ii. Products......................................................................................................... 44
a. Reliability Ratings & Nutrition Labels........................................................ 44
b. Misinformation Fingerprints........................................................................ 48
c. NewsGuard for Advertising.......................................................................... 49
d. Social Impact Reports.................................................................................. 50
iii. U.S. Government Funding............................................................................. 51
a. GEC/DOD Split Award................................................................................ 51
b. Standalone GEC Award................................................................................ 53
2
c. DOD SBIR Award........................................................................................ 54
iv. Partnerships.................................................................................................... 57
V. Recommendations...................................................................................................... 60
1. Investigation into Other GEC Awards................................................................... 60
2. Audit of GEC Award Records and State Records Writ Large............................... 60
3. SBIR Program Reforms......................................................................................... 61
4. Oversight of the NED............................................................................................ 61
5. Legislation.............................................................................................................. 61
VI. Conclusion................................................................................................................. 62
VII. Appendix.................................................................................................................... 63
3
I. Executive Summary
The House Committee on Small Business (Committee or HCSB) is charged with investigating all
“problems of all types of small business” under House Rule X. As part of this responsibility, the Committee
is investigating the Federal funding and promotion of small entities who purposely interfere with the ability of
small domestic businesses to compete online because of their lawful speech, as well as tech start-ups and other
small businesses with products used to surveil and ultimately suppress lawful speech. Over the course of the
Committee’s 14-month investigation, the Committee found:
Finding 1: The Federal government has fueled a censorship ecosystem impacting not only individuals’
First Amendment rights, but the ability of certain small businesses to compete online.
Finding 2: The Federal government has funded, developed, and promoted entities that aim to
demonetize news and information outlets because of their lawful speech, impacting
domestic businesses’ operations, reputation, customer reach, and revenue.
Finding 3: The Global Engagement Center (GEC), an interagency body housed within the U.S.
Department of State (State), circumvented its strict international mandate by funding,
developing, then promoting tech start-ups and other small businesses in the disinformation
detection space to private sector entities with domestic censorship capabilities.
Finding 4: The National Endowment for Democracy (NED), a private non-profit funded almost entirely
by Congressional appropriations, violated its international restrictions by collaborating with
fact-checking entities in assessing domestic press businesses’ admission to a credibility
organization.
Throughout this investigation, the Committee obtained nonpublic GEC and NED award information for
direct and indirect grant recipients. State repeatedly slow rolled Congressional document requests, disregarded
prioritized information, and provided incomplete and inadequate productions. This led the Committee to issue a
subpoena on June 13, 2024. Despite the fact the Committee subpoenaed documents which it had been requesting
for more than 14 months, State said it would take approximately 21 months from the date of the subpoena
to produce these documents in full—around March 2026. In the absence of compliance with the subpoena,
the Committee now remits this interim report on its present findings until such a time where the subpoenaed
information is provided.
4
II. Introduction
Small businesses are the backbone of the U.S. economy, making up more than 99 percent of all domestic
businesses and employing nearly half of the American workforce.1 In order to thrive in today’s market, small
businesses must be freely able to compete online to earn revenue and grow. They leverage social media to reach
new audiences, communicate with customers, market products, earn revenue through platform monetization
capabilities, and build their reputations. Two-sided marketplace platforms are essential for online sales. Hosting
advertisements on a business’ platform is a significant revenue source; conversely, placing advertisements on
other online platforms is a major lever for customer growth. Audiences are reached and web traffic driven through
internet search engines. News and information outlets publish content on the internet and rely less on traditional
paper copies for circulation, using web browsers and social media platforms to reach readers and earn revenue.2
A foundational principle of American markets is that a business will be able to operate without unreasonable
interference from the government so long as they obey the law. However, as extensive investigative reporting
and Congressional investigations show, the Federal government worked with the private sector extensively in
recent years to remove or suppress certain disfavored speech from internet platforms, at the expense of fact,
impacting the ability of businesses purveying that speech to use those services to compete. This collaboration of
public and private entities, including self-proclaimed “fact-checking” organizations, think-tanks, tech companies,
universities, wealthy foundations, and government agencies, has become known as the Censorship-Industrial
Complex (CIC). The reality of the CIC was revealed when Elon Musk purchased Twitter (now X) and “lifted the
Silicon Curtain” to allow investigative journalists to read internal emails between Twitter, government officials,
and government partners. The resulting series of reports are known as the Twitter Files.
Various Federal agencies and their proxies coordinated with social media companies to have disfavored
speech moderated, diminishing impacted businesses’ ability to use those platforms.3 Tech companies changed their
Terms of Service (TOS), the rules which dictate the ability to use each platform, as a direct result of government
pressure, thereby expanding the content and accounts subject to moderation.4 Two-sided marketplace platforms
changed their TOS at the behest of the Biden-Harris White House, interfering with impacted businesses’ sales
revenue.5
If the private companies were making these content moderation decisions on their own volition, without
government interference, that is their right. They are not government entities and are not subject to the same
First Amendment constraints; only the government, and its proxies, are barred from imposing viewpoint-based
1
Frequently Asked Questions About Small Business, U.S. Small Bus. Admin., Office of Advocacy (July 2024).
2
Jacob Liedke & Luxuan Wang, News Platform Fact Sheets, Pew Research Center (Nov. 15, 2023) (finding only 37 percent of U.S.
adults often or sometimes get news from print publications).
3
Several Federal government bodies, including the Biden-Harris White House, the FBI, the GEC, and the Cybersecurity and Infra-
structure Security Agency at the Department of Homeland Security, worked with social media companies to have speech removed
from their platforms. Matt Taibbi, Twitter, the FBI Subsidiary, SubStack (Apr. 13, 2023) originally released on Twitter (now X), Matt
Taibbi (@mtaibbi) (Dec. 16, 2022, 4:00 PM) available at https://x.com/mtaibbi/status/1603857534737072128?lang=en; Matt Taibbi,
New Knowledge, the Global Engagement Center, and State-Sponsored Blacklists, SubStack (Apr. 13, 2023) originally released on
Twitter (now X), Matt Taibbi (@mtaibbi) (Mar. 2, 2023, 12:00 PM) available at https://x.com/mtaibbi/status/1631338650901389322?l
ang=en; H. Comm. on the Judiciary and the Select Subcomm. on the Weaponization of the Federal Gov., The Censorship-In-
dustrial Complex: How Top Biden White House Officials Coerced Big Tech to Censor Americans, True Information, and
Critics of the Biden Administration (May 1, 2024).
4
Facebook (Meta), YouTube, and Amazon, changed their TOS as a direct result of pressure from the Biden-Harris Administration,
enabling increased suppression of speech and removal of content. H. Comm. on the Judiciary and the Select Subcomm. on the
Weaponization of the Federal Gov., The Censorship-Industrial Complex: How Top Biden White House Officials Coerced
Big Tech To Censor Americans, True Information, and Critics of the Biden Administration (May 1, 2024).
5
H. Comm. on the Judiciary and the Select Subcomm. on the Weaponization of the Federal Gov., The Censorship-Industri-
al Complex: How Top Biden White House Officials Coerced Big Tech To Censor Americans, True Information, and Critics
of the Biden Administration (May 1, 2024).
5
censorship restrictions.6 That much of the impacted speech was moderated under mis-, dis-, or malinformation
(collectively, MDM) provisions in the name of safety does not matter- the government cannot coerce third parties to
censor lawful speech.7 It is inconsequential whether the government believes that speech is false, as constitutional
protection does not turn on upon the truth, popularity, or social utility of the ideas and beliefs which are offered.8
The government may not even abridge the freedom of speech, let alone censor it.9
Whether the aforementioned activity rises to unconstitutional violations of the First Amendment is
currently before the courts. The preliminary injunction against the government for alleged social media censorship
in Murthy v. Missouri was reversed and remanded by the Supreme Court (SCOTUS) for the plaintiffs’ failure to
establish Article III standing and the decision therefore did not reach the merits of the case, much to the chagrin
of the dissenting Justices (“The Court, however, shirks that duty and thus permits the successful campaign of
coercion in this case to stand as an attractive model for future officials who want to control what the people say,
hear, and think;” “It was blatantly unconstitutional, and the country may come to regret the Court’s failure to say
so”).10 Mark Zuckerberg has since admitted that the Biden-Harris Administration repeatedly pressured Meta to
censor content on its platforms, a central theme of Murthy.11
In addition to coordination to remove specific content, the Federal government also assisted the private
sector in detecting alleged MDM for moderation. The GEC, though subject to a strict international mandate,
sourced, developed, then platformed and promoted MDM-detection tools directly to the private sector, including
to tech platforms with the ability to moderate domestic speech and impact domestic business operations. The
GEC also worked with foreign governments with strict internet speech laws (including the United Kingdom and
countries in the European Union (EU)) and Federal agencies with no such international restrictions to test these
products.
Among the hundreds of companies promoted by the GEC were press credibility-rating organizations
whose main function is to demonetize news and information outlets purveying speech purported to be MDM by
leveraging partnerships with the advertising industry. These credibility ratings involve the application of subjective
criteria, susceptible to the assessor’s worldview. Ratings are determined not only by what can be proven factually
correct or incorrect, but also by whether narratives are deemed harmful, divisive, or sensational. Certain outlets
are given deference in assessment over others. This is why press ‘trust’ rating systems are inherently flawed- they
are subject to the partisan lens of the assessor.
The purpose of the First Amendment’s Free Speech Clause is to foster “an uninhibited marketplace of
ideas,” testing the “truth” of various ideas in the competition of the market.12 The Federal government cannot
support organizations that interfere with domestic press operations based on whether they are considered
trustworthy or whether they adhere to certain viewpoints. Government abridgement of viewpoint-based speech
6
Police Dep’t v. Mosley, 408 U.S. 92, 95 (1972) (“The First Amendment means that government has no power to restrict expression
because of its message, its ideas, its subject matter, or its content.”).
7
While the definitions of these terms vary slightly depending on source, the general meaning of each is as follows: (1) Misinfor-
mation: false information conveyed without the intent to harm or mislead; (2) Disinformation: false information conveyed with the
intent to harm or mislead; and (3) Malinformation: true information framed in a way that is meant to harm or mislead. By mid-2021
YouTube had removed more than one million videos for alleged COVID-19 misinformation, while Facebook had taken down more
than 18 million pieces of similar content. As blame for the results of the 2016 U.S. Presidential election was laid at the feet of MDM
on social media platforms, it was not just health they sought to protect; there was immense pressure put on these companies to have
and enforce MDM policies to “save democracy” as supposedly threatened by the election of former President Trump. Neal Mohan,
Perspective: Tackling Misinformation on YouTube, YouTube Official Blog (Aug. 25, 2021); Murthy, et al. v. Missouri, et al., 603
U.S. 46 (2024) (Alito, J., dissenting).
8
N.Y. Times Co. v. Sullivan, 376 U.S. 254, 271-72, 84 S. Ct. 710, 721 (1964); see also United States. v. Alvarez, 567 U.S. 709, 718
(2012) (plurality opinion) (“Some false statements are inevitable if there is to be an open and vigorous expression of views in public
and private conversation.”).
9
First Amendment, U.S. Const. amend. I.
10
Murthy, et al. v. Missouri, et al., 603 U.S. 38 (2024) (Alito, J., dissenting).
11
Letter from Mark Zuckerberg, Founder, Chairman, & CEO, Meta Platforms, Inc., to Jim Jordan, Chairman, H. Comm. on the Judi-
ciary (Aug. 26, 2024).
12
Red Lion Broadcasting Co., v. F.C.C., 89 S. Ct. 1794, 1806 (1969).
6
and the press is impermissible regardless of which political party it offends.
The internet age has seen the rise of independent media outlets, the vast majority of which are small
businesses. They must be free to compete online without government interference. At a time when legacy media
is losing trust and audiences by parroting partisan narratives, ironically whether or not that narrative is factually
correct, it is more important than ever that independent media outlets be allowed to compete in the free marketplace
of ideas. However, as award records and awardee operations show, the scales are tipped in favor of outlets which
express certain partisan narratives rather than holding the government accountable.
7
III. Committee Investigation
The Committee opened this investigation following receipt of constituent complaints about partisan ‘fact-
checking’ organizations interfering with their ability to compete online. This was concurrent with investigative
reporting that uncovered the GEC awarding these companies taxpayer dollars.13 Given the domestic small business
impacts, the flurry of reporting on Federal funds and resources being improperly used to target American speech,
and the GEC’s strictly international mandate, the Committee sent an initial request for award records to the GEC
on June 7, 2023. Subsequent requests for additional documentation were made over the course of one year, during
which the Committee gave significant leeway in time and scope to State. Despite these accommodations, only two
heavily redacted lists of awardees were produced, with none of the requested award application, risk assessment,
or contract information.
This fractional production, in addition to State’s significant delays, responsive failures, inattention to
detail, and failure to perform simple due diligence to resolve issues, led the Committee to subpoena State for
the information on June 13, 2024. Despite the subpoena, to date, the vast majority of requested documents have
not been produced. State has informed Committee staff that the subpoena will take an inexcusably long time to
fulfill - approximately 21 months. State has had 14 months to comply with Committee requests and should have
been working on fulfillment prior to the issuance of the subpoena. Further, State and the GEC are subject to
several overlapping investigations which would have caused them to already compile relevant information. The
aforementioned failures are all indicative of a wider problem experienced by other committees- that the Biden-
Harris Administration does not properly adhere to Congressional oversight.
Thanks to extensive work across this space, the Committee was able to learn from the findings of other
committees and investigative journalists when stonewalled by the Biden-Harris Administration. Over the course
of this investigation, the Committee reviewed the following: Federal award applications, contracts, and progress
reports; Federal agency communications, meeting notes, and reports; Office of the Inspector General (OIG)
reports; Congressional committee reports, hearings, and Statements for the Record; Freedom of Information
Act (FOIA) productions; private company reports, press releases, and tax documents; news articles, independent
investigative reports, interviews, court filings and rulings, and statutes. The approximate total pages reviewed is
6,185. The interviews and hearings total approximately 40 hours.
13
Gabe Kaminsky, Disinformation Inc: Meet the groups hauling in cash to secretly blacklist conservative news, Wash. Examiner (Feb.
9, 2023).
8
IV. Small Business: Instruments & Casualties of the
Censorship-Industrial Complex
The Federal government is a primary component of the CIC, and it is this nexus which concerns the
Committee. Investigative journalists and other Congressional committees have examined the Federal government’s
involvement in the CIC in depth, especially pertaining to speech on social media. The GEC is not the only bad
actor in this space, but it is the only Federal entity known to this Committee: (1) whose purpose is countering
foreign disinformation that threatens the United States, (2) that coordinates directly with the private sector to
combat said disinformation, (3) that administers taxpayer dollars to small private entities that interfere with the
ability of American businesses to compete online because of their lawful speech, and (4) that gives platforms to
tech start-ups/small businesses in the surveillance space and promotes their capabilities to private companies,
Federal agencies, and foreign governments that are heavily involved in the censorship apparatus.
Much of the GEC’s work is reputable on its face. Building “resilience to violent extremist propaganda and
ideology” is a worthy cause when the target audience is those susceptible to the influence of ISIS, for example.14
An award purpose of diminishing “the influence of ISIS” and decreasing “its allure in the eyes of potential recruits
and sympathizers” is squarely within the GEC’s mission and Congressional mandate, which is: “[t]o direct, lead,
synchronize, integrate, and coordinate efforts of the Federal Government to recognize, understand, expose,
and counter foreign state and foreign non-state propaganda and disinformation efforts aimed at undermining
or influencing the policies, security, or stability of the United States, its allies, and partner nations.”15 Congress
explicitly included a restriction with that mandate, that “[n]one of the funds authorized […] shall be used for
purposes other than countering foreign propaganda and misinformation that threatens United States national
security.”16
A review of recent direct and indirect awards administered by the GEC in furtherance of its mission
demonstrates that, intentionally or not, the GEC has had impacts on domestic business operations. Speech
surveillance tools owned by domestic small businesses were promoted to the private sector, Federal agencies,
and foreign governments with no international activity restrictions. Federal funds were awarded to grow products
of small entities whose main operations focus on domestic businesses that operate to demonetize businesses
based on lawful speech. There are some awardees who tout ideological beliefs as fact, and work to diminish the
reputation of American businesses if their speech negates those narratives. The GEC should be aware of this, as
they are required to do extensive vetting of each awardee in advance of administering funds and performed similar
due diligence prior to promoting and testing each disinformation detection tool. It is stated verbatim in some
award applications that applicants’ main operations were focused on the speech of businesses based in the U.S.,
and the GEC chose to fund them nonetheless.
14
State Production to H. Comm. on Small Bus. (Dec. 3, 2023); On file with the Comm.
15
Id.; John S. McCain National Defense Authorization Act for Fiscal Year 2019, Pub. L. No. 115-232, § 1284, 132 Stat. 1636, 2076
(2018); Mission & Vision, Global Engagement Center, U.S. Dep’t of State (last visited Jul. 30, 2024).
16
National Defense Authorization Act for Fiscal Year 2017, Pub. L. No. 114-328, § 1287, 130 Stat. at 2548 (2016).
9
1. Global Engagement Center Origin & Mission Expansion
In the wake of the September 11th, 2001 terrorist attacks, there was an interest by policy makers to address
a central question: “why do they hate us?”17 First turning to State’s public diplomacy staff for answers, consensus
was eventually reached that, while public diplomacy was essential to winning the war on terror, the government
was “fighting a 21st century communications war using 20th century tools and platforms.” This void led to the
establishment of the GEC’s precursor entities.
The U.S. Advisory Committee on Public Diplomacy issued a report in May 2024 that narrates the mission
creep of the GEC and its predecessors from the original focus on countering specific terrorist organizations post
9/11, to various global non-state violent extremist threats, to disinformation from state actors like Russia, China,
and Iran. Expansion in resources and private sector coordination are also detailed, as well as hinderances and
priority changes between administrations. It does not cover any activity in the Biden-Harris Administration save
for one sentence on headcount in FY 2021. The report notes that “many historical records […] have not been
preserved, making it difficult to find key documents related to the [GEC and its precursors] and to reconcile
conflicting accounts of events.”18
In January 2016, the Obama White House announced the: (1) establishment of an interagency task force
for countering domestic violent extremism; (2) creation of the GEC; and (3) appointment of Michael Lumpkin as
the new GEC Special Envoy and Coordinator, who brought with him strong ties to the military and intelligence
communities.19 Lumpkin compared the launch of the GEC to “the establishment of the Office of Strategic Services
(the predecessor of the Central Intelligence Agency (CIA)) during World War II” during the GEC’s inaugural staff
meeting.20
Executive Order (EO) 13721 was signed into law on March 14, 2016, officially establishing the GEC
and defining its interagency role. To drive this home, EO 13721 also created a multi-agency Steering Committee
composed of senior representatives of Federal agencies whose work was relevant to the GEC. Chaired by the
Under Secretary of State for Public Diplomacy, members included a senior official from: the Department of
Defense (DOD), the Department of Justice (DOJ), the Department of Homeland Security (DHS), the Department
of Treasury, the Small Business Administration (SBA), the National Counterterrorism Center, the Joint Chiefs of
Staff, the Counterterrorism Center of the CIA, the Broadcasting Board of Governors, and the U.S. Agency for
International Development (USAID).21 It is not clear as to what the Steering Committee looks like in its present
form or what the SBA’s role is as SBA representatives supposedly do not know the answer to that question.22
A significant mission expansion followed with the passage of the FY 2017 National Defense Authorization
Act (NDAA), signed into law by President Obama on December 23, 2016, which: (1) grew the GEC’s authority,
resources, and mandate to support a whole-of-government approach “to expose and counter foreign disinformation
operations and proactively advance fact-based narratives that support U.S. allies and interests;” (2) broadened the
GEC’s threat focus beyond violent extremism to include state actors; (3) gave the GEC grant making authority;
and (4) established a fund to build a network of private sector actors.23
In 2018 sizeable funds were transferred from the DOD to the GEC for initiatives to counter propaganda
and disinformation from foreign nations.24 These included the creation of the Information Access Fund (IAF)
17
The Global Engagement Center: A Historical Overview 2001-2021, U.S. Advisory Comm’n on Public Diplomacy, 11 (May 2024).
18
Id. at 6.
19
Id. at 27.
20
Id.
21
Id at 28.
22
Email from George Holman, Jr., Associate Administrator, Office of Congressional & Leg. Affairs, U.S. Small Bus. Admin., to Steph-
anie Chambless, Senior Investigative Counsel, H. Comm. on Small Bus. (Aug. 28, 2024).
23
National Defense Authorization Act for Fiscal Year 2017, Pub. L. No. 114-328, § 1287, 130 Stat. 2000, 2546- 48 (2016); The Global
Engagement Center: A Historical Overview 2001-2021, U.S. Advisory Comm’n on Public Diplomacy, 30 (May 2024).
24
10
which allowed for the provision of grants to civil society groups, media content providers, nongovernmental
organizations (NGOs), Federally funded research and development (R&D) centers, private companies, and
academia to advance their work in the counter-disinformation and propaganda space.25
In the FY 2019 NDAA, the GEC’s scope was again broadened beyond “lead, synchronize, and coordinate,”
to also “direct” and “integrate” government efforts to counter foreign disinformation and propaganda.26 The
GEC was now aimed at “much more than undermining national security interests” as directed by the original
legislation.27
Though not explicitly stated in the report, this broadened scope accompanied the expansion of the GEC’s
approach from “whole-of-government” to “whole-of-society.” This is reflected in the below talking points
from a symposium hosted by the Atlantic Council’s Digital Forensic Research Lab (DFRLab). This mirrors the
categorization other CIC entities have used when justifying their cooperation with various sectors to promote or
remove certain speech.
28
The two-decades long growth in mission and threats demonstrates why the GEC is a useful tool for
coordinating efforts within the Federal government to detect, understand, and counter dangerous propaganda
circulating internationally. However, somewhere along the GEC’s mission expansion, its methodologies changed
from using social media platforms to create and spread counterpropaganda materials, to directing public opinion
by trying to get social media platforms to suppress content. It is this distinction that is the issue; the government
should not be using its position of power to influence the moderation of lawful speech.
2. The Technology Engagement Team & Private Sector Engagement

The Technology Engagement Division, also known as the Technology Engagement Team (TET), is the
GEC functional unit that focuses on the technological aspects of disinformation and is responsible for working with
social media companies and the tech industry as a whole. The very limited set of the TET’s communication with
the private sector reviewed by this Committee included briefings from tech platforms on their MDM moderation
activity and GEC-directed submission of content the GEC believed violated platform TOS. This included specific
pieces of content and accounts belonging to both individuals and businesses, such as international media outlets
they determined to be purveyors of disinformation.
25
Id. at 33.
26
Id. at 36.
27
Id.; John S. McCain National Defense Authorization Act for Fiscal Year 2019, Pub. L. No. 115-232, § 1284, 132 Stat.1636, 2076
(2018).
28
Global Engagement Center notes for Atlantic Council Digital Forensic Research Lab 360/STRATCOM Roundtable Speech (date
unknown); On file with the Comm.
11
The TET has had “regular ongoing engagements with major [tech] companies (including Microsoft,
Twitter, Facebook/Meta, and Google/YouTube) since 2019.”29 In December 2019, the GEC “established a Silicon
Valley representative to facilitate public-private coordination and broker constructive engagements between the
U.S government and the tech sector, academia, and research.”30 An internal GEC slide deck titled “Tech Sector
Outreach, Company Headlines, and TE Program Highlights” identifies specific tech companies with which the
GEC was in communication, including executive-level points of contact at each company, news highlights, and
additional organizational information. Companies featured included Adobe, Alphabet/Google, Meta (Facebook/
Instagram), Microsoft/LinkedIn, and Twitter (now X).
Communication in the limited records reviewed by the Committee was especially prevalent with Meta and
Twitter, though the GEC’s relationship with both companies evidently waxed and waned between administrations.
As shown by internal emails in the Twitter Files, Twitter was hesitant to work with the GEC during the Trump
Administration (compared to their more established relationships with the FBI and other Federal agencies)
because, among other reasons, it was “more political.”
31
By the time the Biden-Harris Administration had settled in, the TET held monthly meetings with Meta,
with frequent email communication. With Twitter, they held quarterly meetings from at least early 2021, also
with frequent email communication. The TET entered discussions with Twitter for an API developer account in
January 2021 and applied on March 16, 2021, which was evidently rejected by Twitter, then reconsidered less
than a week later following direct outreach by Daniel Kimmage, then Acting Coordinator of the GEC.32
The TET’s relationship with social media companies was so successful that USAID consulted the TET for
tips on how to work with tech companies to get them to be amenable to their content moderation wishes:
29
Email from Global Engagement Center Tech. Engagement Team to USAID (Dec. 15, 2021) (on file with the Comm.).
30
TET Newsletter: The Paperwork Brief, Vol. 2 Ed. 5 (Sep. 3, 2021) (on file with the Comm.).
31
Matt Taibbi, Twitter and the FBI “Belly Button,” SubStack (Apr. 13, 2023) originally released on Twitter (now X), Matt Taibbi (@
mtaibbi) (Jan. 3, 2023, 4:54 PM) available at https://x.com/mtaibbi/status/1610394197730725889?lang=en.
32
Email from Daniel Kimmage, Acting Coordinator, Global Engagement Center, to Twitter (Mar. 19, 2021) (on file with the Comm.).
12
33
3. Awards
The GEC awards produced to the Committee are summarized below, though it is clear some subawards
were not produced. Additionally, single redactions can represent multiple subawardees, and State did not produce
unredacted copies despite the Committee’s multiple requests.
Fiscal
Direct Awards Subawards
Year
2018 39 (37 cooperative agreements, 2 project grants) 48 known, 7 unknown redactions
2019 9 (all cooperative agreements) 9 known, 4 unknown redactions
8 (7 cooperative agreements, 1 direct to
2020 5 known, 4 unknown redactions
individual)
2021 14 (all cooperative agreements) 11 known, 2 unknown redactions
2022 14 awards (all cooperative agreements) 5 known, 5 unknown redactions
2023 24 awards (all cooperative agreements) 5 known, 18 unknown redactions
2024 No information provided No information provided
Of the limited award records reviewed by the Committee, there are many which have direct and downstream
domestic impacts and whose implementation brings the GEC’s compliance with its international mandate into
question. One such award was to the Institute of War & Peace Reporting (IWPR), which administered a subaward
to the Poynter Institute for Media Studies, Inc. (Poynter). This subaward was used to convince international news
outlets to join Poynter’s International Fact-Checking Network’s (IFCN) Code of Principles, which is a group of
news and information organizations that abide by a certain set of qualities denoting journalistic standards.34 To
33
Email from USAID to Global Engagement Center Tech. Engagement Team (Dec. 22, 2021) (on file with the Comm.).
34
The Commitments of the Code of Principles, International Fact-Checking Network, Poynter (last visited Jul. 22, 2024); Countering
Disinformation and Propaganda: Lessons From Practitioners, The Institute for War & Peace Reporting for the Global Engagement
Center, U.S. Dep’t of State (Jun. 16, 2022) (on file with the Comm.); Email from Baybars Örsek, Director, International Fact-Check-
ing Network, to Tech. Engagement Team, Global Engagement Center, U.S. Dep’t of State (Sep. 17, 2020) (on file with the Comm.).
13
join the Code of Principles, organizations must apply and be accepted by the IFCN’s assessors.35 It appears that
in implementing this award, GEC staff was added to a Google Group email in which IFCN assessors (including
representatives from the NED and fact-checking organizations Snopes, Full Fact, and Poynter/IFCN) critiqued
applicants, including domestic businesses such as The Daily Caller and its fact-checking organization.36
The limited communications viewed by this Committee did not show the GEC actively participating in
the conversations of whether to admit applicants to the Code of Principles, only receiving pertinent emails. There
were, however, emails from NED staff, using NED email addresses, opining on The Daily Caller’s credibility.37
38
It is not appropriate, given their strict international constraints, that the GEC or the NED would belong to
a cohort that gatekeeps domestic press companies from belonging to a private credibility organization. As set forth
in its Articles of Incorporation and the National Endowment for Democracy Act, it is a violation of the NED’s
mandate to operate domestically, and therefore to interfere with the operations of domestic press.39
There are several other GEC awards which have domestic business and specifically domestic press
impacts. One such award was to a small domestic business and included deliverables to seek out and promote
internet surveillance products to GEC partner-entities for their use. These partners are not subject to international
restrictions; the private sector partners in the tech industry moderate domestic speech and impact the ability of
domestic small businesses to compete online.
A. Park Advisors
Park Capital Investment Group, LLC (Park Advisors) was a domestic small business founded in 2014 for
which minimal entity information is available. Its GEC award contracts list an address of a house in a residential
neighborhood in Virginia. It is apparently no longer operational as its website has no information other than a
picture of trees.40
In FY 2018, Park Advisors received a GEC cooperative agreement award of more than $6 million.
Deliverables included drafting research reports and assisting the GEC in developing a Testbed methodology,
the Disinfo Cloud platform, a Tech Demo Series, three international Tech Challenges, and a “diverse network of
35
The Commitments of the Code of Principles, International Fact-Checking Network, Poynter (last visited Jul. 22, 2024).
36
Email from Barbara Örsek, Director, International Fact-Checking Network, to [email protected] (Aug. 31, 2022)
(on file with the Comm.); Email from David Mikkelson, Snopes, to [email protected] (Feb. 22, 2021) (on file with the
Comm.).
37
Email from Dean Jackson, National Endowment for Democracy, to [email protected] (Feb. 22, 2021) (on file with
the Comm.).
38
Id.
39
Founding Statement of Principles and Objectives, 1984, National Endowment for Democracy (last visited Aug. 22, 2024); 22 C.F.R.
§ 67 (current as of May 5, 2023); National Endowment for Democracy Act, 22 U.S.C. § 4411.
40
Homepage, Park Advisors (last visited Jul. 19, 2024) https://www.park-advisors.com/.
14
relevant stakeholders.”41
42
In executing its Statement of Work, Park Advisors distributed subawards to several companies, most of
which are small businesses.
Objective Award Purpose Subawards Subaward Amount

Alignment
Atlantic Council $150,000
Babel Street $250,000
To test and engineer novel DROG ADTAC $250,000
technological solutions NewsGuard $50,000
Expose foreign state –hybridization,
through combination, Trend Micro Check $175,000
and foreign non-state applications oforexisting
other
Cyabra $75,000
actors’ tactics and technologies as required by AIfluence $83,000
efforts that spread the GEC – to the problems
propaganda and Sea Monster $80,000
of foreign propaganda and
disinformation to Congo Check $83,000
disinformation, and rapidly
foreign audiences. make those technologies InVid/WeVerify $100,000
available for use by the GEC
and its partners. Institute for Strategic $50,000
Dialogue
Global Disinformation $100,000
Index
43
i. Testbed Methodology & Disinfo Cloud
The main Park Advisors deliverable was to design a Testbed initiative which allowed participants to learn
about and test disinformation detection products in a sandbox environment. The GEC leveraged Park Advisors to
source technologies, largely from small domestic businesses, to add to the sandbox for its partners to test. These
partners– foreign governments, Federal agencies, and the private sector– are not subject to the same international
restrictions as the GEC. The private sector tech partners testing these products moderate not only foreign speech
41
Park Advisors worked with the TET to develop two reports: (1) “Weapons of Mass Distraction: Foreign State-Sponsored Disin-
formation in the Digital Age,” a review of the human and technological vulnerabilities to propaganda and disinformation with an
overview of countries known for such operations; and (2) “Fanning the Flames: Foreign State-Sponsored Disinformation in the Time
of COVID,” which examined how the pandemic and resulting “infodemic” was exploited by hostile state actors and detailed the re-
sponses by governments, online platforms, and civil society. Both reports were distributed to government, academia, think tanks, and
the private sector; Christina Nemr, The Disinfo Cloud Digest, Disinfo Cloud (Dec. 21, 2021); Quarterly Performance Report, Park
Capital Investment Group LLC, dba Park Advisors (Q4 2021) (on file with the Comm.).
42
Quarterly Performance Report, Park Capital Investment Group LLC, dba Park Advisors (Q4 2021) (on file with the Comm.).
43
State Production to H. Comm. on Small Bus., 2 (Mar. 28, 2024) (on file with the Comm.).
15
on their platforms, but also domestic speech and domestic businesses’ use of their products. There was no firewall
in place to ensure that Federal resources were not being used to develop and promote technologies that would
have domestic impacts.
A three-stage methodology was used for assessing and testing technologies, with Stage 1 as evaluation,
Stage 2 as assessments, and Stage 3 for testing.44 Each Stage had nine criteria against which to test the tool and
concluded with a report.45
46
In total, 33 tools reached the operational testing stage with 25 participant offices, including from the GEC,
DOD, embassies, and foreign governments.47 Eight of the tests resulted in longer implementation, while others
resulted in non-GEC participants expressing interest in longer term use pending funding availability.48
The Testbed’s priority was “identifying challenges against which to test technologies,” which this Committee
interprets to mean ‘looking for use cases.’ In conversations with businesses who had products of interest for the
Testbed, to get their buy-in to participate, the most common questions concerned “whether the Testbed would
lead to longer-term contracts,” and “whether companies should expend resources and time engaging in a process
that might not yield such contracts.”49 This led the GEC to recognize the need to “ensure that the Testbed process
first identified offices, agencies, or organizations with relevant use cases and with the necessary budget to spend
if the outputs of the test fill a persistent gap.”50 This conclusion by the GEC suggests that the goal of testing these
technologies was not just for the use of the GEC and may therefore have sought solutions for entities without a
strict international mandate.
The Testbed technologies and their results from the analysis stages were hosted on a searchable repository
called Disinfo Cloud, developed using $1,179,000 of the Park Advisors award.51 The Disinfo Cloud dashboard,
pictured below, displayed the number of technologies on the Testbed, the number of technologies at each stage of
assessment, and the types of technologies they represented.52
44
45
2.09 MadSci Weaponized Information: Technology Engagement team & Disinfo Cloud – Alexis Frisbie & Christina Nemr, Global
Engagement Center, U.S. Department of State, Mad Scientist Weaponized Information Virtual Conference, YouTube (Jul. 21,
2020) available at https://www.youtube.com/watch?v=YoeHq5gX0dA.
46
Id.
47
48
Id.
49
Id.
50
Id.
51
2023.02.14-GEC-GDI-BLACKLIST, 3 (Mar. 2, 2023) (on file with the Comm.).
52
Defeat Disinfo, U.S. Dep’t of State (last visited Aug. 1, 2024).
16
53
Featured products were promoted to Disinfo Cloud’s users, which included Federal agencies, foreign
governments, academia, the private sector (including social media companies), journalists, and civil society
organizations.54
55
There were both private password-protected features of the platform and information on a public-facing
landing page. Categories of disinformation detection tools in the database were listed on the public-facing site, and
users of the password-protected features were able to test those technologies, which totaled 366 tools by project
completion. Hosted technology included: social listening tools, adtech, manipulated information assessment
tools, dark web monitoring, crowd-sourced content assessment and web annotation, blockchain-based media
53
Id.
54
Disinfo Cloud flyer, Tech. Engagement Team, Global Engagement Center (Jul. 15, 2020).
55
Defeat Disinfo, U.S. Dep’t of State (last visited Jul. 13, 2024).
17
authentication, fact-checking, gamified education, and, ironically, internet “censorship circumvention” products.56
A report issued just prior to project close details the following statistics as of December 10, 2021:
a. 365 hosted technologies;

b. 1,883 password-protected users, with the majority from State and DOD;
c. users from 36 foreign governments; and
d. 30,814 individuals who visited Disinfo Cloud’s public-facing platform.57
58
56
Tools Overview, Disinfo Cloud, (Dec. 25, 2022) available at https://web.archive.org/web/20220125194018/https://disinfocloud.
com/tools-overview/.
57
End of Year Accomplishments for CY21, Disinfo Cloud, Global Engagement Center (Dec. 16, 2021) (on file with the Comm.).
58
Id.
18
59
While the vast majority of Testbed companies and products are unknown to the Committee as the subpoena
has not been fulfilled, most of the known companies offer products that surveil social media and other online
speech, often using AI/machine learning, internet wide. Their work is not restricted to international speech and
some of these tools are used by social media companies to track disinformation on their platforms and make
content moderation decisions.
Disinfo Cloud’s blog, branded The Disinfo Cloud Digest, had a “Spotlight” series which showcased the
different technologies on the platform, putting further Federal government support behind them.60 Among the
technologies in the Spotlight series was NewsGuard, discussed further in this report, and Logically.AI, a tool
for monitoring the online media landscape for the spread of potentially harmful narratives, which is not focused
solely on international speech.61 It is alleged in State of Texas, et al. v. U.S. Department of State, et al. that the GEC
featuring Logically.AI in its Spotlight series demonstrates harm to domestic press operations.62 Three months
after Logically.AI was featured in the Spotlight series, Facebook (Meta) retained Logically.AI as a fact-checking
partner.63 Logically.AI explained that when it “rates a piece of content as false, Facebook will significantly reduce
its distribution so that fewer people see it, apply a warning label to let people know that the content has been rated
false, and notify people who try to share it.”64 Logically.AI has implemented multiple negative assessments of the
59
Id.
60
61
The Disinfo Cloud Digest, Disinfo Cloud (Apr. 6, 2021); Homepage, Logically.AI (last visited Jul. 21, 2024).
62
State of Texas, et. al v. U.S. Dept. of State, et al., No. 6:23-cv-00609-JDK, 21 (E.D. Tex. Apr. 10, 2024).
63
State of Texas, et. al v. U.S. Dept. of State, et al., No. 6:23-cv-00609-JDK, 21 (E.D. Tex. Apr. 10, 2024); Stephen Farrell, Logically
Enters UK Fact-Checking Partnership with Facebook, Insider Media (Jul. 16, 2021).
64
Stephen Farrell, Logically Enters UK Fact-Checking Partnership with Facebook, Insider Media (Jul. 16, 2021).
19
domestic press plaintiffs’ content.65
Disinfo Cloud also leveraged Twitter (now X) to promote many of these tools and in several instances,
shared product capabilities outside international speech applications. In the cases of promoting NewsGuard and
the Global Disinformation Index (GDI), this included the ability to withhold American dollars from businesses
sharing disfavored speech.
66 67
The Disinfo Cloud website (www.disinfocloud.com, which did not use a ‘.gov’ top-level domain, despite
being a government-funded platform) is no longer in service as Disinfo Cloud was evidently taken over by a
private party independent of the GEC after the Park Advisors award expired.68 However, a GEC successor called
“Defeat Disinfo” was in planning as of October 2022:
69
ii. Tech Demo Series
Disinfo Cloud served as a way to identify potential technology for the GEC’s Tech Demo Series, a
biweekly forum used to “identify and advocate” for disinformation detection tools to GEC partners.70 Each
Tech Demo agenda began with opening remarks by the GEC, followed by the businesses’ opening remarks, a
presentation, a question-and-answer session, and ended with time for discussion. Participants of the Tech Demos
were encouraged to “sign up [for Disinfo Cloud] to identify and explore technologies under assessment,” thereby
65
State of Texas, et. al v. U.S. Dept. of State, et al., No. 6:23-cv-00609-JDK, 16 (E.D. Tex. Apr. 10, 2024).
66
Disinfo Cloud (@disinfocloud) reposting The Global Disinformation Index (@DisinfoIndex) (Twitter (now X) (Jul. 8, 2020, 6:46
AM) available at https://x.com/disinfoindex/status/1280815562818564096?s=42&t=u8hEK0zejJhIrj8PBhzuqw.
67
Disinfo Cloud (@disinfocloud) reposting NewsGuard (@NewsGuardRating), Twitter (now X) (Mar. 30, 2021, 12:39 PM) available
at https://x.com/newsguardrating/status/1376937058216857606?s=42&t=u8hEK0zejJhIrj8PBhzuqw.
68
Christina Nemr, The Disinfo Cloud Digest, Disinfo Cloud (Dec. 21, 2021); AI-Driven Disinformation Intelligence Platform Black-
bird.AI Announces Jim Reynolds as VP of Sales to Increase Market Reach and Build Strategic Alliances Globally, Business Wire
(Mar. 22, 2022).
69
Email to USARMY, from Counter Disinformation Technology Advisor, Tech. Engagement Division, Global Engagement Center,
U.S. Dep’t of State (Oct. 24, 2022) (on file with the Comm.).
70
TET Slides, Global Engagement Center, 11 (date unknown), available at https://community.apan.org/cfs-file/__key/docpreview-
s/00-00-17-45-79/2.09-TET-Slides-Unclassified-July-17.pdf.
20
further promoting the use of the various speech-policing technologies.”71
More than 100 technologies participated in the GEC’s Tech Demo Series, including psychological
resilience tools, blockchain-based information validation tech, crowdsourced information vetting, and social
media monitoring products.72 Below is a sample of the domestic small businesses of which the Committee is
currently aware whose products were featured in the Tech Demo Series and operate to surveil speech or biometric
data.73 The corresponding information is to the best of the Committee’s knowledge and reflects the companies and
products at the time they were featured in the Tech Demo and is not a current reflection of any updated product
features or entity information (i.e. if the business no longer qualifies as small by SBA standards).
Date of Company Product Product Description

Tech Demo
Blends AI technology with a “24/7 newsroom
of experienced journalists” to combine social
11/20/2019 Factal Factal
media discovery with “trusted verification” in
one platform.
Applies AI to Open-Source Intelligence of
publicly available information in online and
social media. Globally monitors subtle and
Quantitative emerging trends, identifies anomalies in
Scientific
05/19/2021 OpenLine discussions, identifies misinformation and
Solutions, disinformation, and identifies their sources.
LLC (QS-2) Collects text-based data from a variety of social
media sources, including Twitter, Facebook,
and Reddit.74
Video signal processing software that extracts
vital signs, micro-expressions, and other
physiological and biometric data and video
captured on mobile phones and other consumer-
grade video producing devices. Interprets
Presage eSSESSMENT indicators of emotional context, projected
06/30/2021 Security, Inc. and concealed expressions, and physiological
responses to stimuli. Processes millions of
videos from global, publicly available sources.
A previous use case took data from Brown
University to predict COVID-19 risk levels on
a county level in the U.S.75
Pendulum Tracks, measures, and analyzes narratives, sub-
07/14/2021 Pendulum Intelligence narratives, and underlying content from over 20
Platform sources including social media.76
Protagonist.io Identifies, quantifies, and tracks narratives
07/28/2021 Protagonist in media conversations, at scale. Was under
Narrative contract with DOD/Air Force and DARPA.77
Analytics
71
Invitation to Tech Demo 4.15 – Protagonist.io (Jul. 27, 2021) (on file with the Comm.).
72
Past Events, Events – Technology Engagement Division, U.S. Dep’t of State (last visited Aug. 1, 2024).
73
Other Tech Demos featured the following entities and products: Mythos/Carnegie Mellon (Nov. 6, 2019); Mind Over Media (Nov.
20, 2019); KungFu.ai & Data.World (Jun. 24, 2020); Recorded Future (Aug. 22, 2021); LookingGlass (Aug. 24, 2021); Vannevar Labs
(May 25, 2022); Compendium Technologies (Jun. 15, 2022); Premise (Jul. 27, 2022); and Adobe/Content Authenticity Initiative (Aug.
13, 2022).
74
Invitation to Tech Demo 4.10 – QS-2 (May 11, 2021) (on file with the Comm.).
75
Invitation to Tech Demo 4.13 – Presage (Jun. 23, 2021) (on file with the Comm.).
76
Tech Demo 4.14 – Pendulum (Jul. 16, 2021) available at https://www.youtube.com/watch?v=lONJz00XDQ0.
77
Invitation to Tech Demo 4.15 – Protagonist (Jul. 28, 2021) (on file with the Comm.).
21
Social listening technology that provides social
analytics for images, video, and text. Analyzes
memes, GIFs, and videos. Automates narrative
09/22/2021 ViralMoment ViralMoment research on visual platforms like TikTok,
detects and measures viral moments as they are
beginning to emerge before they have reached
critical mass.78
Uses AI to identify the most influential “actors”
Voyager
Analytics and in any cumulative or newly gathered dataset
10/20/2021 Voyager Labs
VoyagerCheck and delves
and into connections as well as related
hidden activities and sponsors.
Small businesses whose products were featured but are not domestic included at least the following:
Date of Company Product Product Description

Tech Demo
A centralized hub for collecting, processing,
and analyzing large volumes of online
media and disseminating insights through
09/08/2021 Geollect InfoSight a dashboard. Integrates datasets from the
physical, informational, and cognitive
dimensions using geospatial contextualization.
Content verification plug-in. Analyzes social
11/3/2021 WeVerify InVid/WeVerify media and web content for disinformation,
misleading and fabricated content.
Uses social listening and sentiment analysis
to derive insights from which to construct
behavior change campaigns that address target
11/30/2022 AIfluence AIfluence issues. Fights disinformation by leveraging
local influencers. Attendees included: DOD,
DOJ, FBI, DHS, CISA, GSA, USAID, and
USAGM.
In sum, the Testbed methodology, the Disinfo Cloud platform, and the Tech Demo Series allowed the
GEC and its partners (including private sector companies with domestic speech censorship operations and foreign
governments with strict internet speech laws) to source and test surveillance and disinformation detection tools.
The GEC marketed and promoted these tools, including by providing Federal funding for some, thereby furthering
the CIC.
iii. International Tech Challenges
The GEC leveraged Park Advisors and Disinfo Cloud resources to host international Tech Challenges to
identify and support companies and products addressing disinformation and propaganda. Applicants and awardees
were largely small businesses/tech start-ups. Prize money was distributed through subawards from Park Advisors’
FY 2018 grant. Through these challenges, Park Advisors identified over 110 tools and technologies, resulting in
eight awardees that used their funding to expand their capabilities in support of the GEC and “other end users,
including civil society.”79
The U.S.-Taiwan Tech Challenge was held in February 2020 and focused on companies working in the
78
Invitation to Tech Demo 4.19 – Viral Moment (Sep. 29, 2021) (on file with the Comm.).
79
22
Asia Pacific.80 It was held in partnership with the American Institute in Taiwan and the Institute for Information
Industry. Awards were distributed to two winners: Trend Micro Check ($175,000) and Cyabra ($75,000).81 Cyabra
was also featured in the Tech Demo Series on April 29, 2020, and is used to protect against “brand reputation
risks, disinformation, and election threats,” and has been retained by companies like Amazon and Disney to scan
social media for content that could negatively impact brand image.”82
The U.S.-Africa Tech Challenge took place in April-May 2021 and sought to address COVID-19 vaccine
hesitancy and disinformation. Among the collaborators were Twitter, USAID, the EU, Ushahidi, Informa, and
the Intergovernmental Authority on Development.83 The three winners were Sea Monster ($100,000), AIfluence
($100,000), and Congo Check ($50,000). The awardees implemented two “digital engagement and behavior
change” campaigns with the primary goal of “increasing positive sentiment towards COVID-19 vaccines” and
to “drive uptake” of the vaccines.84 The campaigns aimed to increase awareness of vaccine disinformation and
disseminate information through community-based influencers and civil society.85 The GEC’s takeaway from
this challenge was that they could leverage influencers to drive behavior campaigns regarding sentiment to the
COVID-19 vaccine.86
An applicant for this challenge, Code for Africa and its subsidiary CivicSignal, submitted its TrustList
product for consideration. TrustList was built with the GDI (discussed in the next section), using the GDI’s
methodology.87 The application stated that Code for Africa/CivicSignal was targeting disinformation in Africa
because “Russia is seeding disinformation amongst African audiences as a backdoor to influence African
Americans and other constituencies in the U.S.”88 This boomerang method of targeting foreign speech because it
impacts Americans is prohibited by the GEC’s international mandate. While Code for Africa/TrustList did not win
the U.S.-Africa Tech Challenge, they have received multiple GEC awards, and likely did not change the reasoning
for its operations between awards applications.89
The U.S.-Paris Tech Challenge was held in June and September of 2021 and was organized in collaboration
with the U.S. Embassy in Paris, the Atlantic Council’s DFRLab, the Cybersecurity and Infrastructure Security
Agency (CISA) within DHS, the North Atlantic Treaty Organization (NATO), the U.K. Department for Digital,
Culture, Media and Sport, the Office of the Tech Ambassador of Denmark, and the Digital Communication
Network.90
The three winners of the U.S.-Paris Tech Challenge were: InVid/WeVerify ($100,000), the Institute for
Strategic Dialogue ($50,000), and the GDI ($100,000). The InVid/WeVerify and GDI subawards were both used
for product development, while the Institute for Strategic Dialogue subaward was used to detect suspected state-
manipulation of Wikipedia, including by interviewing members of the Wikipedia community.91
80
U.S.-Taiwan Tech Challenge, Disinfo Cloud (Jun. 21, 2022) available at https://web.archive.org/web/20220621011352mp_/https://
disinfocloud.com/taiwan-tech-challenge.
81
About Us, Tech. Engagement Team, U.S. Dept. of State Archive (last visited Jul. 20, 2024).
82
About Us, Cyabra (last visited Jul. 30, 2024).
83
U.S.-Africa Tech Challenge, Disinfo Cloud (Oct. 25, 2021) available at https://web.archive.org/web/20211025133937mp_/https://
disinfocloud.com/africa-tech-challenge.
84
Report: Global Engagement Center on RT and Sputnik’s Role in Russia’s Disinformation and Propaganda Ecosystem, Global En-
gagement Center (2021); End of Year Accomplishments for CY21, Tech Challenges (on file with the Comm.).
85
Report: Global Engagement Center on RT and Sputnik’s Role in Russia’s Disinformation and Propaganda Ecosystem, Global En-
gagement Center (2021) (on file with the Comm.).
86
87
Proposal, TrustList @ CivicSignal, Code for Africa (Feb. 28, 2021) (on file with the Comm.).
88
Id.
89
State production to H. Comm. on Small Bus. (Dec. 3, 2023); State production to H. Comm. on Small Bus. (Mar. 28, 2024); On file
with the Comm; Call for applications for newsrooms to help strengthen their Check Desks, Code for Africa (Jan. 26, 2024) avail-
able at https://opportunities.codeforafrica.org/2024/01/26/call-for-applications-for-newsrooms-to-help-strengthen-their-checkdesks/.
90
Paperwork Brief Vol. 2, Ed. 5, Technology Engagement Team, Global Engagement Center, U.S. Dep’t of State (Sep. 3, 2021); End
of the Year Accomplishments for CY21, Tech Challenges (on file with the Comm.).
91
End of the Year Accomplishments for CY21, Tech Challenges (on file with the Comm.).
23
The winners of the U.S.-Paris Tech Challenge were promoted in GEC newsletters and in meetings with the
private sector, including during an October 2021 meeting with YouTube/Google.92
B. Global Disinformation Index
The GDI received at least four Federally funded awards; one directly from the U.S. government through
the U.S.-Paris Tech Challenge, and three from the NED (which was created by Congress and is largely funded
by Congressional appropriations from State), totaling nearly $1 million.93 The GDI is indirectly supported by the
Federal government through continued funding Code for Africa/CivicSignal.
As summarized in its 2020 application for NED funding, the GDI aims “to disrupt, defund and down-rank
disinformation sites, and [the GDI] work[s] collectively with governments, business and civil society to achieve
it.”94
i. About the GDI
The GDI was founded in late 2018 as a U.K. and U.S. based non-profit organization.95 While the original
U.S. 501(c)(3) organization, the AN Foundation, is now defunct, the GDI currently operates through three small
legal entities:
a. Disinformation Index Ltd.: a U.K. private company limited by guarantee;

b. Disinformation Index, Inc.: a U.S. 501(c)(3); and
c. Global Disinformation Index gUG: a nonprofit entrepreneurial company in Germany, analogous to
a limited liability corporation for nonprofit purposes.96
By leveraging partnerships within the advertising industry, the GDI operates to stop the spread of what it
considers to be disinformation by choking off revenue from news and information outlets that have the highest
determined ‘risk’ of spreading said disinformation. This also impacts the business’ reach and reputation. If these
businesses are not able to earn revenue, and their reputations are significantly tarnished, business operations will
suffer, their reach and ability to produce content ultimately diminished. Some outlets targeted by the GDI report
earning just two to six percent of the ad revenue expected for outlets with comparable audience sizes as a result.97
The GDI believes that a main reason for the spread of disinformation is that information/news websites,
particularly those that are low-quality, low-traffic, and quickly constructed, write about certain narratives to drive
traffic to their business; less so that the authors believe those positions or that they are worthy of discussion.98 The
GDI estimates the online advertising industry amounts to $385 billion globally, with nearly $250 million in online
ad revenue earned by ‘disinformation sites’ each year.99 By removing this financial incentive, the GDI believes it
will stop outlets from sharing this information and conform to the GDI’s preferred narratives in order to get better
ratings—harming the viability of businesses that do not.
This curtails speech the GDI does not agree with, because it is not just false information that the GDI labels
as disinformation. Outlets are scored using subjective application of arbitrary criteria, and outlets expressing
92
Internal notes from meeting with Google, Global Engagement Center (Oct. 15, 2021) (on file with the Comm.).
93
The second GDI NED award is a cost (and possibly time) extension of the initial award.
94
Application Documents – Proposal Narrative, National Endowment for Democracy Grant 2020-10474 (2020-1116), Global Disin-
formation Index, 9 (2020) (on file with the Comm.).
95
Id.
96
About, Global Disinformation Index (last visited Jul 20, 2024).
97
Freddie Sayers, Inside the disinformation industry: A government-sponsored agency is censoring journalism, UnHerd (Apr. 17,
2024).
98
Cutting the Funding of Disinformation: The Ad-Tech Solution, Global Disinformation Index (May 1, 2019).
99
Id.
24
speech that does not align with those subjective determinations are subject to the GDI’s interference.
When the GDI was established, it defined ‘disinformation’ as “deliberately false content, designed to
deceive.”100 This is analogous to traditional definitions. However, by 2021, the GDI had expanded its definition of
‘disinformation’ to mean: “adversarial narratives, which are intentionally misleading; financially or ideologically
motivated; and/or, aimed at fostering long-term social, political or economic conflict; and which create a risk of
harm by undermining trust in science or targeting at-risk individuals or institutions.”101
The GDI’s definition of “institutions” includes “the current scientific or medical consensus.” It is vital not
only for the sake of innovation to be able to challenge the current consensus, but also for the ability to resolve
complex issues through public discourse. This problem is demonstrated by the GDI’s position on media outlets
giving credence to the COVID-19 lab-leak theory: “[c]utting off ads to these fringe sites and their outer networks
is the first action needed.”102 This theory on COVID-19 origins is now widely accepted as credible, including as
the “most likely” cause by multiple Federal agencies, yet the GDI had labeled it as disinformation and attacked
the credibility and earning capacity of news and media businesses discussing it.103
The GDI justified the expansion of the ‘disinformation’ definition by saying it allowed their analysis to
include speech that is “harmful” or “divisive.”104 These determinations are subjective and it is clear from the
GDI’s output that it is through a partisan lens. Much of the speech the GDI labels as disinformation under this
definition reflects conservative viewpoints. For example, as the GDI considers speech questioning the efficacy and
safety of the COVID-19 vaccines and opposition to illegal immigration to be adversarial narratives, media outlets
discussing related stances are therefore labeled as purveyors of disinformation under the GDI’s definition.105 This
is not restricted to speech expressing these narratives, but includes the framing of the position; if an outlet uses
the term ‘illegal alien’ rather than ‘undocumented immigrant’ (or the then-current subjectively designated as
politically correct terminology), the GDI will label it as disinformation (under the claim that it creates a risk of
harm by targeting at-risk individuals).106
This demonstrates the central flaw with press-rating organizations, and why they cannot receive Federal
funding or support; speech credibility assessments are subject to the biases of the assessor. It is impossible to rate
the credibility of press objectively using subjective metrics such as whether language is “sensational,” but also
when partisan ideologies are touted as fact. There is always a partisan tilt that manifests when labeling certain
outlets as more trustworthy than others.
The GDI’s politicized interpretation of fact is further evidenced by labeling media outlets asserting
biological differences between men and women, namely, that men cannot become women, as a reason to assign
a poor rating and interrupt their funding. This is demonstrated in its communication with British outlet UnHerd
about its rating:
“Our team re-reviewed the domain, the rating will not change as it continues to have anti-LGBTQI+ narratives…
The site authors have been called out for being anti-trans. Kathleen Stock is acknowledged as a ‘prominent
gender-critical’ feminist.”107
100
Global Disinformation Index, Web Archive (Jul. 1, 2019) https://web.archive.org/web/20190701204010/https:/disinformationin-
dex.org/.
101
Brief: Disinformation Risk in the United States Online Media Market, Global Disinformation Index (Oct. 21, 2022).
102
Evolution of the Wuhan Lab Conspiracy: The Ad-funded Sites Spreading It, Global Disinformation Index (Apr. 24, 2020).
103
The FBI and the Department of Energy found that the COVID-19 lab leak theory on origins is most likely correct; Anumita Kaur &
Dan Diamond, FBI Director Says Covid–19 “Most Likely” Originated From Lab Incident, The Wash. Post (Feb. 28, 2023); see also
Jeremy Herb & Natasha Bertrand, US Energy Department Assesses Covid–19 Likely Resulted From Lab Leak, Furthering US Intel
Divide Over Virus Origin, CNN (Feb. 27, 2023).
104
Clare Melford, Bankrolling Bigotry, London School of Economics IDEAS Online Event, YouTube (Nov. 10, 2021).
105
Phase II Application, U.S.-Paris Tech Challenge, Global Disinformation Index (2021) (on file with the Comm.).
106
Disinformation as Adversarial Narrative Conflict, Global Disinformation Index (Jun. 22, 2022).
107
2024).
25
In this case, the GDI is using a hyper-progressive lens to portray whether an outlet can be trusted based
on whether they adhere to the GDI’s version of truth. Businesses refuting that truth are hurt financially and
reputationally. UnHerd provides a breakdown of the reality of the GDI’s impacts: UnHerd’s ad agency uses the
Grapeshot platform to automatically select websites on which to place ads. Grapeshot uses the GDI to inform
“brand safety;” if the GDI assigns a low score to a media outlet, very few ads will be placed.108
ii. Products
The GDI’s partisan determinations of MDM are factored into its methodology, which includes manual
and automated ratings of news domains/media outlets by level of risk. The underlying assessment framework is
composed of the “expert-identified disinformation flags” spread across four pillars:
109
Based on the aggregate score across the four pillars, each news domain is assigned a risk-level for
disinformation: low, medium-low, medium, medium-high, or high. The scoring scale for the index ranges from
zero (maximum risk of disinformation) to 100 (minimum risk of disinformation). A domain’s score is the mean of
the points earned across all four of the pillars.110
111
The rating system supposedly does not assess whether a specific news outlet is actually purveying
disinformation, but instead assesses the risk of its exposure to disinformation. The GDI says that this “differentiation
is critical,” but it is not clear what that differentiation actually is, as the GDI’s communications with news outlets
and its own reports reveal specific narratives the GDI uses to justify its ratings.112 The GDI also claims it “does not
determine which news stories are inaccurate, or whether a site should be labelled as a disinformation domain. The
index should not be used in this way or to judge what is true news.”113 As demonstrated in the very same report in
which that quote appears, which lists “accuracy of news stories” as a metric, this claim is not accurate.
108
Id.
109
Rating Disinformation Risk: The GDI Methodology, Global Disinformation Index, 6 (Dec. 2019).
110
Id. at 11.
111
Id. at 12.
112
Id. at 5; Freddie Sayers, Inside the disinformation industry: A government-sponsored agency is censoring journalism, UnHerd (Apr.
17, 2024); Evolution of the Wuhan Lab Conspiracy: The Ad-funded Sites Spreading It, Global Disinformation Index (Apr. 24, 2020).
113
Rating Disinformation Risk: The GDI Methodology, Global Disinformation Index, 5 (Dec. 2019).
26
114
a. Reports
The GDI issues periodic reports including “Disinformation Risk Assessments” on media in various global
regions. One such report on American media outlets was prefaced by a press release on October 21, 2022.115 The
accompanying report was issued on December 16, 2022, ranking dozens of American news media businesses by
the GDI’s determination of “trustworthiness” and “risk.”116 Every business the GDI labeled in the top ten “riskiest”
outlets publish content that demonstrates conservative-leaning or libertarian points of view. Every single one of
the supposedly most trustworthy media organizations consistently touts left-leaning narratives.
117
In justifying these results, the GDI assigns qualities it deems each media outlet to possess, which furthers
the evidence of GDI’s partisan bias in its methodology. One such quality is whether the outlet uses “sensational
language,” which turns on the assessor’s definition of that term. The Federalist, for instance, receives a maximum
risk level with the lowest Content pillar scores in the GDI’s study for this reason.118 In contrast, The Washington
Post was assigned a low risk level as it “largely avoids sensational [...] reporting.” It is clear from looking at recent
headlines from opinion pieces such as “Yes, It’s okay to compare Trump to Hitler. Don’t let me stop you” and “Can
114
Id. at 7.
115
116
Disinformation Risk Assessment: The Online News Market in the United States, Global Disinformation Index (Dec. 16, 2022).
117
118
Id.
27
anyone stop the coming Trump dictatorship?” that The Washington Post does not shy from sensationalism.119
Similarly, HuffPost (formerly Huffington Post), which was awarded a low risk level for its “unbiased content free
from sensational text or visuals,” recently posted an article with this headline: “Supreme Court Gives Joe Biden
The Legal OK To Assassinate Donald Trump.” This is a common theme for both outlets historically, but they
generally align with the GDI’s political positions, so that discrepancy seems to have been overlooked.
The evidence of the GDI’s bias is furthered in its application of other metrics, such as whether an outlet
engages in “negative targeting” and “best journalistic practices.” For example, the GDI justified (Business)
Insider’s low risk rating by saying its content “was largely free of bias, negative targeting or sensationalism,
and the articles used journalistic best practices.”120 However, Insider has been subject to frequent criticism for
exclusion of mitigating evidence in its reporting, inaccurate smear campaigns, and contacting subjects of said
campaigns just before publishing without a reasonable timeframe to refute Insider’s claims.
There are two recent examples of Insider’s targeting and lack of journalistic integrity. The first consisted of
two articles accusing the owner of Barstool Sports of criminal behavior while omitting mitigating evidence. Insider
contacted companies that advertised with Barstool in a clear attempt to get them to sever business relationships.121
The second recent example is Insider’s coverage of Bill Ackman’s wife. Ackman had, days before, garnered
national media attention for criticism of Harvard University’s then-president. In this case, Insider provided the
subject with hours to respond before publishing.122 This led Insider’s parent company to perform a review of
Insider’s work.
Other qualities in GDI’s ratings which are clearly applied subjectively include “lack of bias” and whether
an outlet’s reporting is “neutral.” This can be seen in the GDI’s minimum risk rating of NPR, which the GDI
said demonstrated “some small degree of bias” but “neutral, fact-based content.”123 Compare this to the recent
exposé on NPR that was written by a staff member who had been with the institution for 25 years. According to
the exposé, when stories NPR labeled as disinformation turned out to be credible, NPR reportedly “pretended it
never happened” and performed “no self-reflection.”124 The staff member cited specific examples, such as former
President Trump’s supposed Russia collusion, as disproven by the Mueller report, the legitimacy of Hunter Biden’s
laptop, and the COVID-19 lab leak theory.125 The author went on to say there was “no viewpoint diversity” within
NPR’s staff, and that “an open-minded spirit no longer exists within NPR.” Shortly after the exposé was published,
the author was suspended without pay, then subsequently resigned, stating “I cannot work in a newsroom where I
am disparaged by a new CEO whose divisive views confirm the very problems at NPR I cite in my [...] essay.”126
The referenced new CEO once called the First Amendment a “challenge” and stated that “[o]ur reverence for the
truth might be a distraction getting in the way of finding common ground and getting things done.”127
The New York Times (NYT) earned its exceptional rating “in large part based on a high degree of transparency
all around.”128 It is of note that a defector from the NYT, who wrote a book about her experiences there, said in a
recent interview that the NYT’s ‘disinformation consultants’ working with its in-house disinformation department
119
Mike Goodwin, Yes, it’s okay to compare Trump to Hitler. Don’t let me stop you., The Wash. Post (Dec. 20, 2023); Robert Kagan,
Can anyone stop the coming Trump dictatorship, The Wash. Post (Dec. 26, 2023).
120
121
Dave Portnoy (@stoolpresidente), Twitter (now X) (Nov. 4, 2021, 11:56 PM) available at https://x.com/stoolpresidente/sta-
tus/1456304390566993922; Dave Portnoy (@stoolpresidente), Twitter (now X) (Nov. 6, 2021, 9:12 AM) available at https://x.com/
stoolpresidente/status/1456972693178355723?s=10&t=u8hEK0zejJhIrj8PBhzuqw.
122
Neri Oxman (@NeriOxman), Twitter (now X) (Jan. 4, 2024, 2:35 PM), available at https://twitter.com/NeriOxman/status/17429930
73078947843?lang=en.
123
124
Id.
125
Uri Berliner, I’ve Been at NPR for 25 Years. Here’s How We Lost America’s Trust, The Free Press (Apr. 9, 2024).
126
Emily DeLetter, NPR editor Uri Berliner resigns after essay accusing outlet of liberal bias, USA Today (Apr. 17, 2024).
127
Lindsay Kornick, NPR boss once called the First Amendment a ‘challenge’ and ‘reverence for the truth’ a distraction, FOX News
(Apr. 17, 2024).
128
Disinformation Risk Assessment: The Online News Market in the United States, Global Disinformation Index, 19 (Dec. 16,
2022).
28
consider pro-life views to be disinformation, rather than a political ideology they disagree with.129 This politicized
categorization is not disclosed elsewhere known to this Committee and therefore does not suggest “a high degree
of transparency all around.”
These points are made not to belittle the outlets at issue, but to demonstrate that the GDI purports certain
news media organizations that align with its demonstrated political positions as the epitome of journalistic
integrity and unbiased credibility, and those that do not align as “risky” and “untrustworthy.” Subjective metrics
easily impacted by bias are used to reach these determinations. Outlets that do not adhere to the GDI’s views are
subject to reputational and economic harms unless they revise their content in accordance with the GDI’s wishes.
Even if the bias was less egregious, some degree of subjectivity will always exist when rating trustworthiness or
credibility. That is the central, unavoidable flaw, and why the Federal government cannot support organizations
that rate the press, especially those which aim to demonetize certain businesses based on their speech. It is an
abridgement of their First Amendment protections.
During the House Foreign Affairs Subcommittee on Oversight and Accountability (HFAC) hearing on
March 21, 2024 (notably held regarding State’s lack of responsiveness in HFAC’s investigations), State argued
that the GEC’s GDI funding was awarded before the GDI issued its report ranking American media outlets by
risk, implying that the GEC was not aware of the GDI’s work concerning domestic media. However, as the
GDI’s U.S.-Paris Tech Climate application states several times that the entirety of its work up until that point was
focused on English-speaking North America, the GEC was aware of the GDI’s impact on American businesses. In
fact, one of the judging criteria used during the Tech Challenges was the business’ “track record” and “substantive
experience.”130 The GEC understood the nature of the GDI’s work and gave them taxpayer dollars to expand their
capabilities anyway.
b. Dynamic Exclusion List
The GDI’s “core output,” its Dynamic Exclusion List (DEL), is a continually updated list of news and
media businesses that the GDI purports to publish adversarial narratives/disinformation. Launched in May 2020,
the DEL is a licensable tool used by the GDI’s commercial partners to “block monetization services” (cut off
ad spend) from featured businesses.131 If an outlet is on the DEL, that domain is blocked from users’ online ad
bidding systems, resulting in diverting their clients’ ad spend away from content the GDI disapproves of. This is
essentially a blacklist.
As the DEL is proprietary, this Committee does not have significant insight into which outlets are on the
DEL other than those discussed in GDI’s reports. The DEL is discussed in each of the GDI’s Federal funding
applications and was therefore acceptable to the GEC and the NED.
The GDI has been successful in leveraging the DEL to accomplish its goal of defunding disfavored
businesses; the GDI estimates that between launch and 2022, sites on the DEL have lost $128 million in annual
revenue.132 As previously stated, impacted press report earning only two to six percent of the advertising revenue
expected for outlets with comparable audience sizes after being added to the DEL.133
The GEC promoted what appears to be the DEL in a May 2021 meeting with Zoom Video Communications,
Inc. (Zoom). The notes from that meeting reflect several points of note:
129
Nellie Bowles, The Corruption of the American Newsroom, UnHerd (Jun. 5, 2024, at 2:20); Nellie Bowles, Morning After the
Revolution: Dispatches from the Wring Side of History, Penguin Random House (May 14, 2024).
130
Judging Packet, U.S.-Paris Tech Challenge, Global Engagement Center (Sep. 29-20, 2021) (on file with the Comm.).
131
132
Application Documents – Proposal Narrative, National Endowment for Democracy Grant 2022-1181, Global Disinformation Index,
1 (2022) (on file with the Comm.).
133
2024).
29
1. Zoom staff asked about “lists” that could be shared around “malign actors,” to which the GEC
recommended the GDI and the Hamilton 2.0 dashboard.
2. The GEC promoted its Testbed project, sharing insights into tests with bit.ly across Twitter and
Telegram, and also the potential to utilize synthetic detection technologies to identify synthetic text
in Zoom’s chat function and/or synthetic photos in Zoom avatars or background uploads.
3. The GEC invited Zoom to use its Disinfo Cloud platform to learn more about the disinformation
detection technologies the GEC was working with.134
c. Veracity.ai
Veracity.ai is the GDI’s software-as-a-service (SaaS) product that won Federal funding through the U.S.-
Paris Tech Challenge.135 It was prototyped in 2017 using a grant from the Knight Foundation, then was scaled up
with grants from the U.K. Foreign and Development Office (U.K. FDCO) and the Omidyar Network.136 It is an
AI-driven disinformation risk detection platform that measures disinformation risk across the internet and blocks
monetization from outlets it labels as purveyors of disinformation. Data is combined with machine learning and
human review to identify websites spreading disinformation, which are then added to the DEL.
In its application for the U.S.-Paris Tech Challenge, the GDI stated that end goal of Veracity.ai was to
block monetization from outlets the product determined to be purveyors of disinformation.137 At the time of
application to the Tech Challenge, Veracity.ai covered over 300,000 internet domains and millions of pieces of
content on a weekly basis.138
139
134
Internal notes from meeting with Zoom Video Communications, Inc., Global Engagement Center (May 27, 2021) (on file with the
Comm.) during this meeting, Zoom communicated that while it did not proactively scan their meetings around the world, or meeting
titles (as an example) for keywords, they may do so in the future.
135
136
Id.
137
Id.
138
Id.
139
Id.
30
iii. Funding
The GDI earns revenue from government awards, donations from philanthropic organizations, and license
fees for its DEL and Veracity.ai products.140 The GDI has received U.S. taxpayer dollars from at least two sources:
the GEC (through a subaward from Park Advisors for winning the U.S.-Paris Tech Challenge) and the NED. The
GDI’s work was concentrated in North America at the time of application to both entities, which both the GEC
and the NED were aware of. The GDI leveraged the GEC and NED funding to develop its infrastructure and
expand its work beyond the English language and North America, seemingly in-line with both organizations’
strict international mandates, yet this only spread the GDI’s perception of rated businesses to other audiences and
allowed it to grow its capabilities.
The GDI says the following in its initial NED application, ignoring that they are “the unscrupulous
companies [governments] hire”:
141
They continue with more ironic truths: “[g]overnments are increasingly advancing measures which
threaten the very vitality of an independent and trusted media that they aim to protect.”142
a. GEC Funding
The GDI submitted its application for the U.S.-Paris Tech Challenge through its British entity, Disinformation
Index, Ltd. At the time of application, the GDI had already received funding from the U.K government and U.S.
foundations, with licensing revenues accounting for about 18 percent of its total income.143 It had partnered with a
dozen ad tech companies by that time and was redirecting millions of dollars of ad spend.144 It boasted that over the
previous year, it had cut the number of ad auctions to sites they labeled as disinformation in half.145 The application
notes that at the time of submission, Veracity.ai supported mainly English-language content out of North America,
so the financial impact referred to was mostly felt by English-speaking North American businesses.146
140
About, Global Disinformation Index (last visited Jul. 20, 2024).
141
142
143
144
Id.
145
Id.
146
Id.
31
147
147
Id.
32
148
The GDI reportedly used the GEC funding to “accelerate its scale up.”149 By growing the GDI, the GEC
enabled the GDI to further its work on U.S. media outlets, demonstrated by the fact that the GDI published its
report ranking U.S. media outlets after it received the GEC funding.
The GEC also promoted the GDI through the Disinfo Cloud Twitter account and directly to the private
sector in the TET’s communications with technology companies, including Zoom and Google/YouTube. It is
therefore not only the Federal funding that is at issue, but also promotion and validation with the weight of the
U.S. State Department behind it.
b. NED Funding
The NED awarded a total of $756,923 to the GDI’s two U.S. nonprofit entities: $545,750 composed of
an original award and a cost extension to the now defunct AN Foundation (NED1 & NED2, respectively); and
$211,173 to the active Disinformation Index, Inc. (NED3).
148
Id.
149
The Global Disinformation Index (@DisinfoIndex), Twitter (now X) (Sep. 30, 2021, 12:03 PM) available at https://x.com/disin-
foindex/status/1443607364503015436?s=42&t=u8hEK0zejJhIrj8PBhzuqw.
33
Amount
Acronym NED Grant Number Grantee Name Start Date End Date
Paid
NED1 2020-10474 (2020-1116) AN Foundation 7/1/2020 2/17/2023 $230,000
2020-10474 (2020-1116)
NED2 AN Foundation 7/1/2020 2/17/2023 $315,750
Cost-Extension
Disinformation
NED3 2022-1181 10/1/2022 2/17/2023 $211,173
Index, Inc.
TOTAL $756,923
150
The Grant Agreement for NED1 indicates an original end date of June 30, 2021, presumably extended by
the administration of NED2 with the additional funding. The NED2 Grant Agreement viewed by this Committee
is identical to NED1 (in fact, all documents produced for NED2 are identical to NED1), indicating either a
production error or that no additional documentation was required for the additional funding. The intended end
date of the funding to the AN Foundation is therefore unclear. The intended end date for NED3 was September 30,
2024. The Grant Agreement for NED3 indicates that the intended funding amount for that award was $1,022,931.
151
However, all NED funding was prematurely terminated on February 17, 2023. This coincides with pressure
stemming from investigative reporting about the GDI’s partnerships and activities.152 A statement issued by the
NED to the Washington Examiner confirms that funding was terminated as the GDI’s work on domestic entities
gives the appearance of contradicting the NED’s strictly international mandate:
“As set forth in our Articles of Incorporation and the NED Act, our mandate is to work around
the world and not in the United States. We have strict policies and practices in place so that NED
and the work we fund remains internationally focused, ensuring the Endowment does not become
involved in domestic politics. Recently, we became aware that one of our grantees, the Global
Disinformation Index (GDI), was engaged in an initiative, funded by a different donor, that focused
150
Original table created by the NED with the Acronym column added by the Comm. (on file with the Comm.).
151
Agreement, Grant 2022-1181, Global Disinformation Index and National Endowment for Democracy, 1 (Sep. 25, 2022) (on file
with the Comm.).
152
Gabe Kaminsky, Disinformation Inc: Government-backed organization sent $315,000 to group blacklisting conservative news, The
Wash. Examiner (Feb. 14, 2023).
34
on specific U.S. media outlets. We recognize the important work GDI has done with NED support
in other countries to help preserve the integrity of the information space and counter authoritarian
influence. However, given our commitment to avoid the perception that NED is engaged in any
work domestically, directly or indirectly, we will no longer provide financial support to GDI.”153
The NED grant documents show that the NED was aware of the GDI’s work in the U.S. media market at
the time of funding. In its NED1 application, the GDI also discloses its relationship with the World Federation of
Advertisers (a body that represents roughly 90 percent of global advertising spend—almost one trillion dollars),
the Global Alliance for Responsible Media (GARM), and other advertising bodies that heavily impact domestic
advertising business:154
155
Further, the Program Activities in the NED3 Project Description included the following, which seemingly
would have impacted domestic entities in addition to the international work: (1) expand methodologies and
capabilities to new “surfaces” including video (e.g., YouTube, broadcast, or cable television news), audio (e.g.,
radio broadcasts, podcasts), and apps (e.g., mobile, connected TV); and (2) further develop GDI’s pilot YouTube
capability by implementing the full integration of the YT-DEL (presumably stands for YouTube Dynamic
Exclusion List) channel identification capabilities with its existing Dynamic Exclusion List Quality Assurance
process.156
The initial NED application (NED1) reported existing funding from the U.K. FDCO (committed funding
of £1.4 million), Luminate (committed funding of $1.4 million), Craig Newmark Philanthropies ($120,000)
and Becker Trust ($50,000).157 The GDI’s relationship with the U.K. FDCO was so entrenched that its Head of
Programme, Susan Stern, was listed as a reference in the funding application.158
153
Gabe Kaminsky, Disinformation Inc: State Department-backed group cuts ties with group blacklisting conservative news, The
Wash. Examiner (Feb. 20, 2023).
154
Allum Bokhari, The Advertising Industry’s Deepening Role in Online Censorship, Foundation For Freedom Online (Jul. 31, 2024).
155
156
Agreement, Grant 2022-1181, Global Disinformation Index and National Endowment for Democracy, 3 (Sep. 25, 2022) (on file
with the Comm.).
157
158
Id.
35
By the time the GDI submitted its 2022 application for NED3, it had received the following funding:
159
iv. Partnerships
The GDI leverages its partnerships in the advertising space to achieve its aim of defunding disfavored
outlets. In its application for the U.S.-Paris Tech Challenge, the GDI dictates likely users, audiences, and partners
for its Veracity.ai product.
160
The GDI’s customers at the time of its GEC application included: MediaMath, IAS, Oracle Data Cloud,
Quantcast, Peer39, Omnicom Media Group, Adthrive, Boston Scientific, Flatfile, Infolinks, OpenWeb, and
Microsoft.161 It also listed the French government as a partner, citing its support of the Open Terms Archive.
The GDI displays two dozen partnerships on its website, presumably to which it provides results of its risk
determinations or are product licensees.
159
Application Documents – Proposal Narrative, National Endowment for Democracy Grant 2022-1181, Global Disinformation Index,
14 (2022) (on file with the Comm.).
160
161
Id.
36
162
Notably missing is Microsoft, which was a named partner in its U.S.-Paris Tech Challenge application
and whose advertising company Xander used the GDI.163 Upon inquiry by the Washington Examiner, Microsoft
claimed it was examining the relationship with the GDI and would be suspending services in the interim.164 It is
unclear whether Microsoft followed through.
The GDI lost other partnerships because of its partisan methods, including with the multinational software
company Oracle, whose Vice President for Corporate Communications stated: “[a]fter conducting a review, we
agree with others in the advertising industry that the services we provide marketers must be in full support of free
speech, which is why we are ending our relationship with GDI.”165
The U.K. FDCO (which had awarded approximately $3.2 million (£2.6 million) to the GDI) similarly
severed ties after alarm bells were rung by UnHerd, drawing the attention of the U.K. Business Secretary and ten
members of British Parliament.166 The GDI’s practice of down-ranking outlets if they assert differences in gender
was said to run afoul a protected belief in British law.167
v. GEC-GDI-Blacklist.docx
The Committee obtained a copy of the internal GEC document titled “2023.02.14 GEC-GDI-BLACKLIST.
docx”, which is dated March 2, 2023, despite its file name. It is a “voluntary” press guidance for GEC staff to use
when asked about the GEC’s relationship with the GDI, the Park Advisors award projects, and the GEC’s work
generally. The document instructs readers to inform inquirers that the Park Advisors award is no longer effect, has
been archived, and if asked for any records to redirect to the FOIA process.168
Several claims are made in the document that are called into question by investigative reporting and
internal GEC communications, documents, and requisite procedures. The first is that “the GEC does not and has
never attempted to moderate content on social media platforms” (emphasis added).169 It is a stated goal of the
GEC to work with social media platforms to flag potentially TOS violative speech. This goal was communicated
by the State Department at large to its Bureaus to coordinate with the GEC to do so with global content on U.S.
platforms. That the U.S. government is asking companies to review this content on a voluntary basis does not
162
About, Global Disinformation Index (last visited Jul. 20, 2024).
163
164
Gabe Kaminsky, Disinformation Inc: Microsoft suspends relationship with group blacklisting conservative news, Wash. Examiner
(Feb. 11, 2023).
165
Gabe Kaminsky, Disinformation Inc: Massive corporation Oracle severs ties with conservative blacklist group, The Wash. Examin-
er (Apr. 19, 2023).
166
Alex Farber, MPs oppose funding disinformation ratings agency in blacklisting row, The Times (Apr. 19, 2024); Frederick Attenbor-
ough, Governments Are Beginning to Resist ‘Disinformation Index’, The European Conservative (May 2, 2024); Archie Earle, David
Cameron: Government will no longer fund Global Disinformation Index, UnHerd (May 9, 2024).
167
As decided in the British case Forstater v Centre for Global Development Europe, the belief that sex is biological and immutable
is a “protected philosophical belief” under Britain’s Equality Act 2010. Maya Forstater -v- CGD Europe, Center for Global Develop-
ment, Masood Ahmed, Courts and Tribunals Judiciary (Jul. 6, 2022).
168
2022.02.14.GEC-GDI-Blacklist.docx (on file with the Comm.).
169
Id.
37
mean that the government is not attempting to moderate content.
170
Even the limited emails reviewed by this Committee between the GEC and social media companies show
GEC staff sending specific content to platforms the GEC believed to violate platform TOS. It is obvious that by
doing so, the GEC sought to have this content actioned. TOS dictate what is allowed on a platform; by flagging
violations, the GEC was sending platforms content the GEC did not believe should be on the platform. This is
an attempt to moderate content. Similar to reporting content on platforms from personal accounts, the intent is
to have the content actioned; except this is coming from the State Department of the United States Government.
There is an inherent pressure with these submissions that does not exist when a personal account reports content
as a possible TOS violation.
Further, the below email from USAID shows staff seeking the GEC’s assistance with “content moderation
issues.”
171
The GEC’s caveats that its submissions were for the platform’s “situational awareness” means only that
the government was aiming to shield itself from First Amendment scrutiny.
170
Interagency Action Plan for Ethiopia on Countering Hate Speech, Global Engagement Center, 7 (Nov. 2021) (On file with the
Comm.).
171
Email from USAID to Global Engagement Center Tech. Engagement Team (Dec. 22, 2021) (on file with the Comm.).
38
172
173
172
Email from Tech. Engagement Team, Global Engagement Center, U.S. Dep’t of State, to Neema Guliani and Stacia Cardille, Twit-
ter (Dec. 29, 2020) (on file with the Comm.).
173
Email from Tech. Engagement Team, Global Engagement Center, U.S. Dep’t of State, to Neema Guliani, Todd O’Boyle, and Stacia
Cardille, Twitter (May 10, 2021) (on file with the Comm.).
39
174
The disclaimers were not present on all emails containing content flagged by the GEC.
175
174
Email from Tech. Engagement Team, Global Engagement Center, U.S. Dep’t of State, to Neema Guliani, Todd O’Boyle, and Stacia
Cardille, Twitter (May 10, 2021) (on file with the Comm.).
175
Email from Tech. Engagement Team, Global Engagement Center, U.S. Dep’t of State, to Neema Guliani, Twitter (May 2, 2022) (on
file with the Comm.).
40
The document continues with the claim that “[t]he GEC does not fund programs in the U.S.”.176 This
statement turns on the interpretation of ‘programs.’ If ‘programs’ is interpreted to mean ‘projects’ or ‘initiatives,’
which is more likely, while the GEC does not fund initiatives domestically focused on their face, many of them
do impact the U.S., as discussed throughout this report.
If ‘programs’ is interpreted to mean ‘organizations,’ which is less likely, it is also incorrect as the GEC has
funded many organizations that, to the best of the Committee’s knowledge, are U.S.-based or have offices in the
U.S, such as:
1) Albany Associates International (office in Bethesda, MD);

2) Atlantic Council;
3) Atlantic Council’s DFRLab;
4) Babylon, Inc.;
5) Center for European Policy Analysis;
6) Center for Strategic and International Studies;
7) CNA Corporation;
8) College of William and Mary;
9) Democracy Council of California;
10) DT Institute;
11) East-West Center;
12) E-Collaborative for Civic Education;
13) Equal Access International;
14) Freedom House;
15) GLOBSEC (office in Washington, D.C.);
16) Institute for War and Peace Reporting (office in Washington, D.C.);
17) International Center for Journalists;
18) International Center for Religion & Diplomacy;
19) International Republican Institute;
20) International Research and Exchanges Board;
21) National Democratic Institute for International Affairs;
22) NewsGuard (subaward);
23) Park Capital Investment Group;
24) Poynter Institute (subaward);
25) Project Harmony;
26) RAND Corporation;
27) Sayara International;
28) Search for Common Ground;
29) TechSoup;
30) The Critical Mass, LLC; and
31) Thomson Reuters Foundation.
The document also states that “Park Advisors served as our third-party implementer and administered
the selection of GDI during the U.S.-Paris Tech Challenge in 2021 and issued the sub-award thereafter.”177 This
response is framed as if the GEC was not involved in selection of the U.S.-Paris Tech Challenge winners. The
Director of the TET was one of the judges.178
Further, this suggests Park Advisors independently issued the GDI sub-award. Every award, and subaward,
176
2022.02.14.GEC-GDI-Blacklist.docx (on file with the Comm.).
177
Id.
178
Program Packet, U.S.-Paris Tech Challenge (Sep. 29-30, 2021) (on file with the Comm.); U.S.-Paris Tech Challenge, Atlantic
Council (last visited Aug. 2, 2024) available at https://www.atlanticcouncil.org/event/u-s-paris-tech-challenge/.
41
must be approved by the GEC. As relayed by the GEC grant officers and grant officer representatives during the
April 9, 2024, briefing with Committee staff, subawards are very rarely administered prior to discussion with the
GEC as awardees are required to obtain permission to administer them. Further, each awardee and subawardee is
required to be assessed for various types of risk by GEC staff before granting the award.
Much of the remainder of the document is dedicated to discrediting the Twitter Files. As it was released
on March 2, 2023, it is predated by Twitter Files #1-17, coincidentally (or not) the same day that Twitter Files #17
was published (“New Knowledge, The Global Engagement Center, and State-Sponsored Blacklists”).179
C. NewsGuard
Self-anointed as “The Internet Trust Tool,” NewsGuard is a domestic for-profit business that rates the
credibility of news and information outlets and tells readers and advertisers which outlets they can trust. In addition
to online media outlets, NewsGuard also rates podcasts, blogs, broadcast news programs and television networks.
Since its launch in 2018, NewsGuard has rated more than 35,000 news and information sources, accounting for
95 percent of online engagement with news across nine countries.180
These ratings, combined with NewsGuard’s vast partnerships in the advertising industry, select winners
and losers in the news media space. NewsGuard guides spending and traffic toward approved press and away from
outlets that NewsGuard deems to misalign with its metrics. Low ratings are particularly harmful to small outlets;
their ability to compete is impeded by loss of much needed revenue for growth and a tarnished reputation as they
are portrayed as unreliable, reducing circulation. These harms are remedied and ratings increased by adhering to
NewsGuard’s requirements.
Like with all credibility rating systems that include subjective criteria, it is impossible to achieve objectivity
in the outcomes. While NewsGuard publishes its assessments, offering transparency into its work, this does not
negate that ratings are assigned through an unavoidable partisan lens.
i. About NewsGuard
NewsGuard Technologies, Inc. is a for-profit entity headquartered in New York. Its initial funding of
$6 million was led by Publicis Groupe, with a total of 18 investors.181 While headcount indicates it remains a
small business, as it is a private company the Committee cannot confirm the financials against SBA maximums
for current small business categorization. It was a small business at the time of its Federal awards, at least, as it
qualified for the Small Business Innovation & Research (SBIR) award program in September of 2021.
NewsGuard has a global staff of ‘trained journalists’ and ‘information specialists.’ Its Board of Directors
includes both co-founders/co-CEOs and the Chief Operating Officer of Publicis Groupe.182 NewsGuard also
has an Advisory Board that lends advice and subject-matter expertise; it is a commonsense assumption that the
company’s thematic perception of fact is influenced by these members. NewsGuard asserts its Advisory Board
plays no role in the press ratings or assessments unless otherwise noted and has no role in the governance or
management of the organization.183 On the Advisory Board sits Richard Stengal, the former head of the GEC, Tom
Ridge, the former Secretary of DHS, Anders Fogh Rasmussen, the former head of NATO, and Michael Hayden,
179
Matt Taibbi, New Knowledge, the Global Engagement Center, and State-Sponsored Blacklists, SubStack (Apr. 13, 2023) originally
released on Twitter (now X), Matt Taibbi (@mtaibbi) (Mar. 2, 2023, 12:00 PM) available at https://x.com/mtaibbi/status/16313386509
01389322?lang=en.
180
About NewsGuard, NewsGuard (last visited Jul. 20, 2024).
181
Brill and Crovitz Announce Launch of NewsGuard to Fight Fake News, Publicis Groupe (Mar. 5, 2018).
182
Board of Directors, NewsGuard (last visited Aug. 6, 2024).
183
Advisory Board, NewsGuard (last visited Jul. 19, 2024).
42
a retired four-star general and former head of the CIA and NSA.184
NewsGuard is similar to the GDI in that it operates to stop the spread of disinformation using financial
levers; they aim for the advertising revenue of businesses purported to spread MDM to be “targeted and whittled
away” and for offending outlets to be “systemically defund[ed].”185 However, NewsGuard’s methodology, and its
justification for its ratings, are more transparent than the GDI’s, providing citations with their fact-checking. This
does not mean that its methodology and output are free from bias or error; NewsGuard’s TOS states that it does
not warrant that its services will be error-free, and that there may be inaccuracies in the content or other material
made available through it services.186
NewsGuard also aims to protect ‘brand safety’ by helping blue-chip companies stop placing their ads on
‘harmful’ or ‘untrustworthy’ sites; the examples given include Chinese and Russian disinformation outlets, or
those “peddling quack remedies.”187 However, it is not just these blatantly dangerous outlets that are subject to
NewsGuard’s ratings, but essentially all domestic press. NewsGuard’s impact is also far more significant than the
GDI’s as its partnerships are much more expansive. Since it announced its relationships with State and the DOD,
NewsGuard’s private partnerships have grown enormously, spanning from the highest echelons of advertising
groups to tech companies to educational organizations. This impacts domestic press’ revenue and reach in each of
those industries.
NewsGuard claims to believe “the answer to misinformation is not blocking content or censoring speech,
but instead arming people with information that provides the context and digital literacy skills they need.”188
This statement, which mirrors the foundational beliefs of this nation and of this Committee, does not align with
NewsGuard in practice. NewsGuard leverages “human intelligence” (journalists on staff) to dictate an outlet’s
trustworthiness.189 Those deemed “untrustworthy” are then compiled into “exclusion lists,” with “trustworthy”
sites on “inclusion lists,” which are licensed to advertisers to instruct their ad agencies and ad-tech partners to
keep their programmatic ads off/on these sites.190 This, in practice, impacts business revenue and reputation,
the ability to compete, to retain staff, and generate content. Ratings accompanying links to news articles when
the NewsGuard plugin is installed impact pageviews and reputation. It is a financial and perception-imposing
backdoor to stifling the reach of certain outlets and amplifying others.
NewsGuard goes to great lengths to create the appearance of nonpartisanship and objectivity. Following
prolonged media coverage and Congressional scrutiny on NewsGuard, its methodology, and its links to the
government, NewsGuard’s co-founder/co-CEO Gordon Crovitz has written multiple articles in defense of its
work.191 It is true that NewsGuard gives fairer treatment to conservative-leaning businesses than the GDI. It is also
true that its methodology is more transparent, with less evidence of partisan ideology being touted as fact. That
does not mean that its treatment of the media is equal or that its methodology is free from error; nor does it mean
that the Federal government should be giving taxpayer dollars to a company who operates to demonetize domestic
press on a basis of their lawful speech, tells users which news businesses can be trusted, and profits off licensing
press rating systems and ‘exclusion lists.’ Whether it aims to be objective or not, the government cannot support
such an organization whose operations, by design, interfere with the press in this manner.
184
Id.; Michael Hayden is a signatory of the infamous letter by intelligence officers asserting that Hunter Biden’s laptop was Russian
disinformation (a false claim that was used to ban business accounts like that of the New York Post from social media), yet he sits on
the board of the company that tells users what is true or false and which businesses should be trusted.
185
Social Impact Report 2021, NewsGuard, 1 (Jan. 2022); NewsGuard Expands Service to Australia and New Zealand, Rating News
Sources and Tracking False Narratives; Finds Climate Change Misinformation to be Major Subject of Unreliable Websites, News-
Guard (Mar. 15, 2023).
186
Terms of Service, NewsGuard (last visited Jul. 19, 2024).
187
Gordon Crovitz, Only transparent, apolitical ratings for news publishers can be trusted, The Wash. Examiner (Feb. 13, 2023).
188
Social Impact Report 2021, NewsGuard, 13 (Jan. 2022).
189
About, NewsGuard (last visited Aug. 15, 2024).
190
Matt Skibinski, Special Report: Top brands are sending $2.6 billion to misinformation websites each year, NewsGuard (last visit-
ed Aug. 15, 2024).
191
Gordon Crovitz, Advertisers fear supporting journalism, here’s how to fix that, The. Wash. Examiner (Jul. 19, 2024); Gordon Cro-
vitz, Only transparent, apolitical ratings for news publishers can be trusted, The Wash. Examiner (Feb. 13, 2023).
43
It is not this Committee’s intention or responsibility to comb through all of NewsGuard’s ratings for
inconsistencies in its work. It is NewsGuard’s prerogative and right to assess media as it sees fit and to act upon
those assessments in the marketplace, as it is for all non-government actors. The Constitution guarantees that
the government shall not interfere with the freedom of the press, however; Federal funds and support cannot be
awarded to such an entity. A system that rates the credibility of press is fatally flawed as it is subject to the partisan
lens of the assessor, making the ratings unreliable.
ii. Products
a. Reliability Ratings & Nutrition Labels
Reliability Ratings are NewsGuard’s scores for media and information outlets. These ratings are based on
nine criteria and graded on a scale of zero to 100. Resulting totals dictate the level of caution NewsGuard suggests
when consuming content from each outlet. As the ratings are assigned by outlet, rather than article-by-article, the
rating stays the same regardless of topic or author; the same author can publish the same piece through a different
outlet and the article will be attributed a completely different rating.
1) 100: High Credibility – the outlet adheres to all nine standards of credibility and transparency.
2) 75-99: Generally Credible – the outlet mostly adheres to basic standards of credibility and
transparency.
3) 60 – 74: Credible with Exceptions – the outlet generally maintains basic standards of credibility
and transparency, without significant exceptions.
4) 40 – 59: Proceed with Caution – the outlet is unreliable because it fails to adhere to several
basic journalistic practices.
5) 0 – 39: Proceed with Maximum Caution – this website is unreliable because it severely violates
basic journalistic standards.192
If a web browser has NewsGuard’s software installed, each outlet’s rating will appear next to its articles in
online search engine results, social media posts of the articles, as well as on the articles themselves. If searching
for news on a particular topic, and a user is faced with choosing a 100/100 rated outlet deemed to exhibit “High
Credibility” versus a low rated outlet with a “Proceed with Maximum Caution” warning, it is more likely that user
will choose and trust the content with the higher rating.193 This becomes problematic when high-rated outlets are
not being honest or objective, but the user is told they are trustworthy.
The weighted metrics used to determine each rating include qualifications necessary to earn the points
in each category, which are awarded on a pass/fail basis; an outlet receives either all points associated with each
criterion, or none.194
1) 22 points: does not repeatedly publish false or egregiously misleading content.

2) 18 points: gathers and presents information responsibly.
3) 12.5 points: has effective practices for correcting errors.
4) 12.5 points: handles the difference between news and opinion responsibly.
5) 10 points: avoids deceptive headlines.
6) 7.5 points: website discloses ownership and financing.
7) 7.5 points: clearly labels advertising.
8) 5 points: reveals who’s in charge, including possible conflicts of interest.
9) 5 points: provides the names of content creators, along with either contact or biographical
information.195
192
Website Rating Process and Criteria, NewsGuard (last visited Jul. 19, 2024).
193
NewsGuard’s Online Source Rating Tool: User Experience, Gallup (Jan. 2019).
194
195
Id.
44
NewsGuard explains that while in some instances the pass/fail system may make the ratings less precise,
the alternative of awarding partial points for each is not feasible for the volume of news outlets they rate.196 This
is one area where subjective decision-making and bias seem to enter NewsGuard’s rating process; the threshold of
when a category’s points are forfeit is not clear and appears to differ by outlet. This is demonstrated by reviewing
the ‘Nutrition Label,’ or scoring justification, of each outlet, which includes: (1) the evidence and examples
NewsGuard uses to defend each assessment, (2) any relevant exchanges with the outlet about the score, and (3)
a history of the outlet’s ratings.197 While heavily contributing to NewsGuard’s transparency, it sheds light on
deference given to outlets receiving categorical points despite blatant, repeat violations that are omitted from its
Nutrition Label.
NewsGuard claims to be fighting for the truth to prevail, yet when a high-ranking outlet is misreporting a
story, receives the points in content categories, violations are not cited in its Nutrition Labels, and the user is told
the outlet is trustworthy, NewsGuard is ironically doing the opposite. This is demonstrated with the mainstream
media’s selective framing that former President Trump warned of a ‘bloodbath’ to come if he does not win the
2024 Presidential election, when he was talking about the auto-industry.198 The below are all deceptive headlines
that are not reflected in the respective outlets’ Nutrition Labels. Each of these outlets receives the points for the
“avoids deceptive headlines” category, despite the category requirement that “slightly sensational” headlines not
misrepresent content.
1) The Associated Press (100/100) via NPR (100/100): “Trump says some migrants are ‘not people’
and warns of ‘bloodbath’ if he loses.”199
2) The Associated Press (100/100) via CBS News (90/100): “In Ohio campaign rally, Trump says
there will be a “bloodbath” if he loses November election.”200
3) NBC News (100/100): “Trump says there will be a ‘bloodbath’ if he loses the election.”201
4) NYT (87.5/100): Trump Says Some Migrants Are ‘Not People’ and Predicts a ‘Blood Bath’ if He
Loses;202 Trumps Warning of a ‘Blood Bath’ if He Loses203
5) Politico (100/100): Trump says country faces ‘bloodbath’ if Biden wins in November204
6) The Washington Post (100/100): “The bloodbath Trump promised has already begun.”205
Another example is the recent assassination attempt on former President Trump. The following headlines
are similarly omitted from the high-rated outlets’ Nutrition Labels and the outlets receive the “avoids deceptive
headlines” category points.
1) The Associated Press (95/100) via ABC News (75/100): “Donald Trump escorted off stage by
Secret Service during rally after loud noises ring out in crowd.”206
2) CNN (80/100): “Secret Service rushes Trump off stage after he falls at rally;”207 “Trump injured
in incident at rally.”208
196
Id.
197
Id.
198
Id.; David Emery, Did Trump Say It Will Be a ‘Bloodbath for the Country’ If He Doesn’t Get Elected?, Snopes (Mar. 17, 2024).
199
Trump says some migrants are ‘not people’ and warns of ‘bloodbath’ if he loses, NPR, The Associated Press (Mar. 17, 2024).
200
In Ohio campaign rally, Trump says there will be a “bloodbath” if he loses November election, CBS News, The Associated Press
(Mar. 18, 2024).
201
Emma Barnett & Jillian Frankel, Trump says there will be a ‘bloodbath’ if he loses the election, NBC News (Mar. 16, 2024).
202
Anjali Huynh & Michael Gold, Trump Says Some Migrants Are ‘Not People’ and Predicts a ‘Blood Bath’ if He Loses, The N.Y.
Times (Mar. 18, 2024).
203
Trumps Warning of a ‘Blood Bath’ if He Loses, The N.Y. Times (Mar. 18, 2024).
204
Myah Ward, Trump says country faces ‘bloodbath’ if Biden wins in November, Politico (Mar. 16, 2024). This is despite a March 11,
2024 Politico headline titled “Bloodbath at RNC: Trump team slashes staff at committee.”
205
Dana Milbank, The bloodbath Trump promised has already begun, The Wash. Post (Mar. 22, 2024).
206
Donald Trump escorted off stage by Secret Service during rally after loud noises ring out in crowd, ABC News via The Associated
Press (Jul. 13, 2024).
207
Secret Service rushes Trump off stage after he falls at rally, CNN (Jul. 13, 2024).
208
Trump injured at incident at rally, CNN (Jul. 13, 2024).
45
3) NBC News (100/100): “Secret Service rushes Trump offstage after popping noises heard at his
Pennsylvania rally.”209
4) NewsWeek (100/100): “Trump campaign launches GoFundMe after shooting;”210 “MAGA
Responds With Outrage After Donald Trump Injured at Pennsylvania Rally.”211
5) NYT (87.5/100): “Trump Rushed Off Stage After Chaos at Rally.”212
6) USA Today (100/100): “Trump removed from stage by Secret Service after loud noises startles
former president, crowd.”213
7) The Washington Post (100/100): “Trump escorted away after loud noises at Pa. rally.”214
Relying only on these high-ranked media, readers may have had no idea that there was an assassination
attempt on the former President. They may have thought he set up a fundraiser for himself, rather than the families
and victims at the rally. While NewsGuard seeks to “restore trust in the media,” with many issues they are making
it harder for the truth to come out.
This is a theme with many of the content categories. There are many outlets that receive the points for
“gathers and presents information responsibly,” for instance, despite numerous violations omitted from their
Nutrition Labels. Examples include The New Republic (92.5/100) and The Washington Post (100/100), each
with many headlines exhibiting precisely the type of rhetoric that would reasonably be considered irresponsible,
but evidently fails to register, even though this metric requires outlets to be fair and accurate in reporting and
presenting information.215 That some of these headlines are attached to opinion pieces does not negate their
irresponsibility.
1) The New Republic (92.5/100): “Trump Will Be a Dictator on Day One and Every Day Thereafter;”216
X post: “Behind all of Elon Musk’s bloviating and attention-seeking is a small man who is simply not
very good at anything” linking to article: “ Elon Musk is The New Republic’s 2023 Scoundrel of the
Year – He proved this year that he’s not just evil, he’s deeply stupid, too;”217 Series: “What American
Fascism Would Look Like.”218
(This article series is headed by a picture of former President Trump as Adolf Hitler.)
2) The Washington Post (100/100): “Yes, It’s okay to compare Trump to Hitler. Don’t let me stop
you;”219 “A Trump dictatorship is increasingly inevitable. We should stop pretending;”220 “The Trump
209
Secret Service rushes Trump offstage after popping noises heard at his Pennsylvania rally, NBC News (Jul. 13, 2024).
210
Flynn Nichols, Trump campaign launches GoFundMe after shooting, NewsWeek (Jul. 14, 2024).
211
Jason Lemon, MAGA Responds With Outrage After Donald Trump Injured at Pennsylvania Rally, NewsWeek (Jul.13, 2024).
212
Trump Rushed Off Stage After Chaos at Rally, The N. Y. Times (Jul. 13, 2024).
213
Trump removed from stage by Secret Service after loud noises startles former president, crowd, USAToday (Jul. 13, 2024)
214
Trump escorted away after loud noises at Pa. rally, The Wash. Post (Jul. 13, 2024).
215
216
Matt Ford, Trump Will Be a Dictator On Day One and Every Day Thereafter, The New Republic (Dec. 6, 2023).
217
The New Republic (@newrepublic), Twitter (now X) (Aug. 24, 2024, 7:12 PM) available at https://x.com/newrepublic/sta-
tus/1827484106503901293?s=42; Alex Shepard, Elon Musk Is The New Republic’s 2023 Scoundrel of the Year, The New Republic
(Dec. 27, 2023).
218
Michael Tomasky, et al., (Series) What American Fascism Would Look Like, The New Republic (2024).
219
Mike Goodwin, Yes, it’s okay to compare Trump to Hitler. Don’t let me stop you., The Wash. Post (Dec. 20, 2023).
220
Robert Kagan, A Trump dictatorship is increasingly inevitable. We should stop pretending, The Wash. Post (Nov. 30, 2023).
46
dictatorship: How to stop it;”221 “If Donald Trump became a dictator, who could stop him?;”222 “Can
anyone stop the coming Trump dictatorship?”223
Further, the “handles the difference between news and opinion responsibly” metric requires that opinionated
language generally not be presented as news and that opinion pieces be clearly labeled.224 NewsGuard claims that
its scoring process is designed to ensure its criteria are applied equally and accurately to all sites, irrespective of
an outlet’s political leaning.225 There are many media outlets with top scores, however, that repeatedly exhibit
clear bias and purport opinion as fact without discern but are awarded the points in this category. One blatant
example of this is The Atlantic, which receives a perfect 100/100 score despite consistently touting left-leaning
narratives without clearly differentiating opinion pieces.226 Recent examples include: “Trump Can’t Deal with
Harris’s Success;”227 “Suddenly Trump Looks Older and More Deranged;”228 “A Searing Reminder That Trump
Is Unwell;”229 “Trump Is Suddenly Running Scared;”230 “Trump Should Never Have Had This Platform;”231 “We
Still Don’t Know What to Do With the Endless Stream of Trump Lies;”232 and “The Gunman and the Would-Be
Dictator.”233
This metric also requires that if an outlet has an overall agenda or point of view that it advances by its
choice of the stories it covers or the views it advances that those be clearly disclosed.234 It is no secret that legacy
media paints different political parties in vastly different lights. This is a pattern and why it is no surprise that trust
in the media has declined.235 Conservative figures are consistently portrayed negatively while Democratic figures
receive puff pieces and are not scrutinized in the same manner. It is as dishonest for “The Internet Trust Tool” to
ignore this as it is for the mainstream media to pretend their bias doesn’t exist.
This theme is exhibited in the framing of Democratic campaigns (such as the recent CBS News (90/100)
portrayal of Vice President Harris’ flip-flopping as “moderation” of her policies) and when presenting the same
position stemming from opposing campaigns.236 CBS’s coverage on the policy platform of removing tax from tips
when proposed by former President Trump compared with Vice President Harris exemplifies this:
221
Robert Kagan, The Trump dictatorship: How to stop it, The Wash. Post (Nov. 30, 2023).
222
Amber Phillips & Jillian Banner, If Donald Trump became a dictator, who could stop him?, The Wash. Post (Dec. 18, 2023).
223
Robert Kagan, Can anyone stop the coming Trump dictatorship, The Wash. Post (Dec. 26, 2023).
224
225
Id.
226
Many of these headlines have a ‘Politics’ category header, but do not designate them as opinion.
227
Peter Wehner, Trump Can’t Deal With Harris’s Success, The Atlantic (Aug. 12, 2024).
228
Anne Applebaum, Suddenly Trump Looks Older and More Deranged, The Atlantic (Jul. 22, 2024).
229
Tom Nichols, A Searing Reminder That Trump Is Unwell, The Atlantic (Jul. 19, 2024).
230
David A. Graham, Trump Is Suddenly Running Scared, The Atlantic (Aug. 2, 2024).
231
David Frum, Trump Should Never Have Had This Platform, The Atlantic (Jun. 28, 2024).
232
David A. Graham, We Still Don’t Know What to Do With the Endless Stream of Trump Lies, The Atlantic (July 11, 2024).
233
David Frum, The Gunman and the Would-Be Dictator, The Atlantic (Jul. 14, 2024).
234
235
Megan Brenan, Americans’ Trust In Media Remains Near Record Low, Gallup (Oct. 18, 2022) (finding just 7 percent of Americans
have a “great deal” of trust and confidence in the media, while 28 percent say they do not have very much confidence and 38 percent
have none at all in newspapers, TV, and radio).
236
CBS News (@CBSNews), Here’s how Kamala Harris is moderating some of her more controversial policy stances as a presiden-
tial candidate, Twitter (now X) (Aug. 15, 2024) available at https://x.com/cbsnews/status/1824272367452815724?s=10, linking to
Kathryn Watson, The evolution of Kamala Harris’ stances on single-payer healthcare, fracking, and the Supreme Court, CBS News
(Aug. 15, 2024).
47
237
It is not clear what NewsGuard’s threshold is to remove categorical points. It is not clear how it chooses
which outlets to rate or which to heavily scrutinize over others. It is not clear that the issues they assess in Nutrition
Labels are evenly spread across the political aisle or that, despite their claims, that politics has no impact on their
assessments.238 This is especially true given NewsGuard’s recent questioning of George Washington University
Law School professor Jonathan Turley about his political leanings and why they aren’t disclosed on his blog.239 As
Professor Turley asks in his reporting of the conversation, is this question posed to all outlets NewsGuard rates?240
Drawing no conclusions, it is of note that Professor Turley wrote an article about NewsGuard roughly one week
before he was contacted for rating.
While NewsGuard claims their “work rating news sources and identifying false narratives […] is entirely
independent and free of any outside influence,” it does not seem to be free of inside influence.241 As stated, press
rating systems are inherently flawed because they are subject to the partisan lens of the assessor. That these ratings
have significant impacts on the operation, revenue, reputation, and reach of domestic media businesses drives
home the same conclusion: press ‘trust’ rating organizations should not receive Federal funding or support.
b. Misinformation Fingerprints
In the process of rating outlets, NewsGuard catalogues what it determines to be the most prominent
falsehoods and “misinformation narratives” spreading across these sites, “capturing data and examples about
each hoax that can be used as a “Fingerprint” for the hoax by AI and machine learning tools.”242 Misinformation
Fingerprints is the resulting licensable dataset of these narratives, covering topics such as election fraud and
COVID-19. It is human curated, and so the dataset content and theme (as well as evidence used to argue fact or
falsehood) are subject to biases of NewsGuard’s staff. Launched in 2021, the catalog is continuously updated and
machine-readable, “purpose-built for artificial-intelligence tools to identify the provenance of hoaxes and track
237
CBS News (@cbsnews), Twitter (now X) (Jun. 17, 2024, 4:40 PM) available at https://x.com/cbsnews/sta-
tus/1802803402545467446?s=10; CBS News (@cbsnews), Twitter (now X) (Aug. 12, 2024, 1:30 PM) available at https://x.com/
cbsnews/status/1823049308230324279?s=10.
238
Website Rating Process and Criteria, NewsGuard (last visited Jul. 19, 2024) NewsGuard claims that its scoring process is de-
signed to ensure its criteria are applied equally and accurately to all sites, irrespective of an outlet’s political leaning.
239
Jonathan Turley, A Shield or Sword? A Response to NewsGuard, Res ipsa loquitur – The thing itself speaks (Jul. 29, 2024).
240
Jonathan Turley, The Most Chilling Words Today: I’m from NewsGuard and I am Here to Rate You, Res ipsa loquitur – The thing
itself speaks (Jul. 29, 2024).
241
Email from Gordon Crovitz, NewsGuard, Co-CEO, to Matt Taibbi, RacketNews, Investigative Journalist (Mar. 10, 2023).
242
NewsGuard wins Pentagon-State Department contest for detecting COVID-19 misinformation and disinformation, NewsGuard
(Aug. 18, 2020).
48
the spread of narratives in real time.”243 In practice, this means that Misinformation Fingerprints allow users to
pick a narrative from the database and search websites and social media to see who is using that speech.244
Accompanying each “Fingerprint” is the following data:
1) Example language used to advance the narrative;

2) Variations of the narrative;
3) Evidence supporting why that narrative is false;
4) Source and timing of each narrative’s emergence;
5) Links to examples of the narrative online;
6) Related keywords and hashtags; and
7) A “risk of harm” designation.245
246
By the end of 2021, 762 “false narratives” had been entered into the Misinformation Fingerprints catalog.247
The dataset grew to 1,122 Fingerprints by the end of 2022, and 1,887 by the end of 2023.248
As demonstrated in the U.S. Government Funding section, taxpayer dollars were leveraged to develop the
Misinformation Fingerprints product. The Pentagon and State were the first to use Misinformation Fingerprints to
“counter disinformation,” essentially outsourcing the U.S. government’s perception of fact to NewsGuard.249
c. NewsGuard for Advertising
NewsGuard for Advertising is a product for the advertising industry that uses NewsGuard’s Reliability
Ratings/Nutrition Labels to curate ‘inclusion’ and ‘exclusion’ lists of outlets assigned various levels of
trustworthiness.250 This impacts rated media outlets’ ability to earn advertising revenue; spending is driven away
from outlets on the exclusion lists, which, similar to the GDI’s Dynamic Exclusion List, are essentially blacklists.
243
244
Id. at 36.
245
Misinformation Fingerprints, NewsGuard (last visited Jul. 19, 2024).
246
247
248
Id.; Social Impact Report 2023, NewsGuard, 5 (Jan. 2024).
249
250
Social Impact Report 2020, NewsGuard, 19, (2020); NewsGuard Launches Six New Brand Safety Tiers, Enabling Greater Control
for Advertisers to Support Credible News and Avoid Misinformation, NewsGuard (Dec. 7, 2022).
49
Pre-assembled lists are offered in various tiers depending on what the brands or their advertising partners
are looking for; if they want to “maximize safety” they can choose to exclude all websites NewsGuard deems
to be spreaders of misinformation or conspiracies, but also those which they deem to be credible but have had
significant exceptions in journalistic standards.251 There are pre-filled lists that maximize reach while still avoiding
outlets deemed to spread clearly harmful disinformation; this is subject to what NewsGuard’s analysis determines
to be lower-tier reliable news sites.252 Users of NewsGuard for Advertising are also able to craft custom news
inclusion or exclusion lists “based on their standards,” but if the users are paying for NewsGuard’s services and
NewsGuard is telling them an outlet isn’t trustworthy, it is obviously less likely the advertiser will ignore their
findings and place ads anyway.
Packages are also offered specifically for election misinformation. This includes website, podcast, and
television ratings, as well as access to NewsGuard’s election-focused exclusion list, complete with regularly
scheduled “check-ins” with NewsGuard’s team to discuss specific misinformation narratives.253
NewsGuard’s reach in this space is demonstrated by its vast partnerships. The ad industry is evidently
placing less and less ads on news outlets generally for fear of spiking controversy, but that does not mean that
ads have completely disappeared from the press. The programmatic advertising industry places 15 million ad
impressions per second, every second of the day.254 The average campaign for larger advertisers places these ads
across 44,000 websites.255 This is a huge industry and revenue source in which the scales are no longer even. It
impacts the ability of domestic press to compete online. For small outlets on exclusion lists, it interferes with
much needed revenue for growth.
d. Social Impact Reports
NewsGuard releases annual Social Impact Reports detailing their work. The scale of NewsGuard’s impact
on the news and media industry is informed by these statistics.
By the end of 2021, NewsGuard had rated 7,466 domains covering 95 percent of online engagement.256
Seven million public library patrons used their Nutrition Label product. NewsGuard’s ratings were used as a
factor in what global news aggregators curated, reaching 625 million monthly users.257 This is indicative of
NewsGuard’s impact on the ability for a business to grow and succeed. If NewsGuard deems it untrustworthy, it
is less likely to be included in a news aggregator site using their product.
NewsGuard had reviewed 8,641 domains by the end of 2022.258 In January 2022, NewsGuard secured a
partnership with the American Federation of Teachers, giving access to its Reliability Ratings browser extension
to 1.7 million teachers, their students, and families.259 Students’ perception of who to trust was therefore in the
hands of 35 NewsGuard staff.260 This is furthered by NewsGuard’s partnership with Turnitin, which is used by
over 34 million students.261
By the end of 2023, NewsGuard had rated more than 10,000 websites and more than 35,000 total
publishers spanning websites, social media platforms, podcasts, and television channels.262 Most of these are
251
252
Id.
253
Id. at 17.
254
NewsGuard for Advertising, NewsGuard (last visited Jul. 23, 2024).
255
Id.
256
257
Id. at12.
258
259
Id.
260
261
Turnitin Partners with NewsGuard to Offer Students and Teachers its Media Literacy Tools, NewsGuard (May 4, 2020).
262
Social Impact Report 2023, NewsGuard, 5 (Jan. 2024); About, NewsGuard (last visited Jul. 29, 2024).
50
small businesses, not legacy media outlets. It is hard enough for small and independent press outlets to compete
with mainstream media without the interference of an organization that impedes their reputation, circulation, and
ability to earn advertising revenue, let alone one that has received Federal funding and support.
iii. U.S. Government Funding
NewsGuard has received Federal funding from the GEC and the DOD. The structure of these awards
included licensing fees for the Misinformation Fingerprints database and for NewsGuard to carry out government-
research work that further developed its product. Those initial relationships with significant U.S. government
bodies were then promoted by NewsGuard, and extensive partnerships in the private sector followed.
While NewsGuard presents these transactions purely as licensing agreements for its Misinformation
Fingerprints product (“[t]hese licenses are only for access to our data”), it is fundamental to recipients of DOD
SBIR awards that they engage in research/R&D to meet the needs of the Armed Services.263 NewsGuard also
states in its 2021 Social Impact Report that it used the SBIR award to “further develop” its Misinformation
Fingerprints program.
264
a. GEC/DOD Split Award
Although not disclosed in production to this Committee (as only the below $50,000 Park Advisors
subaward from 2022 is reflected), NewsGuard was issued $25,000 in November 2020 through a Park Advisors
subaward for a four-month project to help the GEC and USCYBERCOM better understand the origins, content,
and spread of certain foreign disinformation campaigns.265
Described as a “pilot” award, this is the first known Federal funding to NewsGuard, which evidently led
to the second Park Advisors/GEC award and the DOD SBIR award. By winning the competition, NewsGuard’s
Misinformation Fingerprints tool was subsequently “sponsored” on the GEC’s Testbed/Disinfo Cloud, and
therefore promoted to the GEC’s partners.266
Reminiscent of the Minority Report’s “pre-crime,” the competition sought a solution to “identifying
hoaxes and misinformation in advance” — what NewsGuard described as “prebunking” of “hoaxes from its
already identified sources of misinformation.”267 While NewsGuard’s press release about winning the award says
263
Email from Gordon Crovitz, NewsGuard, Co-CEO, to Matt Taibbi, RacketNews, Investigative Journalist (Mar. 10, 2023); The SBIR
and STTR Programs, U.S. Small Bus. Admin. (last visited Apr. 17, 2024).
264
265
On file with the Comm.
266
NSIN Challenge – Countering COVID19 Disinformation, National Security Innovation Network (last visited Jul. 21, 2024).
267
(Aug. 18, 2020).
51
the work would be focused on “identifying online sources spreading COVID-19 disinformation,” the affiliated
Statement of Work is focused on disinformation designed to influence the 2020 U.S. elections.268 COVID-19
narratives were presumably folded into this.
269
Although recent discussion of this award suggested only the Misinformation Fingerprints tool was used
for this and other Federal projects, NewsGuard’s press release about winning the competition suggested the
Reliability Ratings/Nutrition Labels products were also involved, at least in the application if not execution of the
award.
270
However, we know from NewsGuard’s description of its products that Reliability Ratings inform the
Misinformation Fingerprints database and that there is not a firewall between the two products.271
268
Id.
269
Contract Agreement, NewsGuard Tech, Inc. and Park Capital Investment Group LLC, 9 (Nov. 2, 2020) (on file with the Comm.).
270
(Aug. 18, 2020).
271
Id.
52
b. Standalone GEC Award
NewsGuard performed additional work for the GEC in 2022 through a second Park Advisors subaward of
$50,000, which included a license to NewsGuard’s Misinformation Fingerprints program.
272
The project was related to “Russian disinformation narratives” circulating in Venezuelan media outlets,
and sought to “test the effect and impact of overlaying NewsGuard’s content accuracy data atop existing content
authenticity data generated through two separate pilots, run by Truepic and Serelay.”273 The $50,000 fee was
evenly split between the two different projects with Truepic and Serelay.274 Truepic and Serelay collaborated
with NewsGuard to “A/B test user perception of trust in media when shown authenticity verification alone
versus authenticity verification paired with NewsGuard reliability verification.”275 While testing was restricted to
international audiences only, it is not clear whether the media reviewed was solely international.276 This question
is furthered by an internal GEC slide deck discussing the results of the project, in which 44 percent of the flagged
websites were hosted in the U.S.
272
Misinformation Fingerprints License Agreement, NewsGuard Tech, Inc., to Park Capital Investment Group, LLC dba Park Advisors
(Jan.10 2022) (on file with the Comm.).
273
Contract Agreement, NewsGuard Tech, Inc. and Park Capital Investment Group LLC, 11 (Dec. 31, 2021) (on file with the Comm.).
274
Id.
275
Id.
276
Id.
53
277
The GEC reportedly no longer subscribes to the Misinformation Fingerprints tool, nor does it have a
present working relationship with NewsGuard to the Committee’s knowledge. The GEC cannot, however, claim
that it was not involved in Park Advisors’ decision to administer the subaward to NewsGuard (as it did with the
GDI); in addition to the requisite review the Federal government must complete for each subaward, the Contract
Agreement states: “Park Advisors and [NewsGuard], with significant input from the GEC, agreed upon a fixed
price amount and scope of services to be delivered as part of this contract” (emphasis added).278
c. DOD SBIR Award
The SBIR program is known as America’s seed fund. As with most government funding opportunities, it
is a highly competitive program; it encourages domestic small businesses to engage in Federal R&D that has the
potential for commercialization. The SBIR program is meant to stimulate high-tech innovation while enforcing
the entrepreneurial spirit essential to this country’s economy and growth. Since 2013, the SBIR program has been
used to administer over 4,500 awards, supporting small businesses in all 50 states.279
Each year, Federal agencies that have extramural R&D budgets exceeding $100 million are required to
allocate a certain percentage of that budget to the SBIR program.280 The SBA directs participating agencies in
program administration by helping them implement the SBIR program, reviews their progress, reports annually to
Congress on its operation, and aggregates agency solicitation announcement information.281 The SBA also directs
program policy for all participating agencies and provides program oversight.282 Considering a recent theme of
SBIR awards has been developing disinformation detection tools, including Meedan and NewsGuard, it is unclear
whether that means the SBA has guided SBIR program policy in this direction.
277
Tech Sector Outreach, Company Headlines, and TE Program Highlights, Global Engagement Center (date unknown) (on file with
the Comm.).
278
Contract Agreement, NewsGuard Tech, Inc. and Park Capital Investment Group LLC, 1 (Dec. 31, 2021) (on file with the Comm.).
279
Eleanor Johnson, NSF 101: America’s Seed Fund, National Science Foundation (Nov. 2, 2023).
280
FAQs, Small Business Innovation Research (SBIR) & Small Business Technology Transfer (STTR), U.S. Small Bus. Admin. (last
visited Jun. 29, 2024) available at https://www.sbir.gov/faqs.
281
Id.
282
Eligibility Requirements, Small Business Innovation Research (SBIR) & Small Business Technology Transfer (STTR), U.S. Small
Bus. Admin. (last visited Jun. 29, 2024) available at https://www.sbir.gov/faqs/eligibility-requirements.
54
There are three Phases, or categories, of SBIR awards:
1) Phase I: establish the technical merit, feasibility, and commercial potential of the proposed Federal
Research /R&D efforts and to determine the quality of performance of the small business awardee
organization prior to providing further Federal support in Phase II. Phase I awards normally do not
exceed $150,000 and have a six-month period of performance.283 However, some agencies offer larger
awards, such as the National Science Foundation, which administers Phase I awards of nearly $300,000
for six to twelve months of work.284
2) Phase II: continues the Federal Research/R&D efforts initiated in Phase I. Funding is based on the
results achieved in Phase I and the scientific and technical merit and commercial potential of the
project proposed in Phase II. Generally, only Phase I awardees are eligible for a Phase II award, but
as demonstrated with the NewsGuard award, there are also Direct to Phase II options. Phase II awards
normally do not exceed $1,000,000 and have a two-year period of performance.285
3) Phase III: The objective, where appropriate, is for the small business to pursue commercialization
objectives resulting from the Phase I and Phase II Federal Research/R&D activities. The SBIR program
does not fund Phase III awards. In some Federal agencies, Phase III may involve follow-on non-SBIR
funded R&D or production contracts for products, processes or services intended for use by the U.S.
Government.286
As NewsGuard’s SBIR award was a Direct to Phase II, the split GEC/DOD Park Advisors subaward
(which NewsGuard described as a “pilot” award) seems to have served as the Phase I prerequisite to Phase II,
despite not being administered under the SBIR program. This is furthered by the government’s description of the
competition’s award, which would provide the winner with a “Government Contracting 101 session and SBIR
crash course.”287
Recipient NewsGuard Technologies, Inc.

FAIN FA8649-21-P-1569
Awarding Agen- Department of Defense
cy
Branch Air Force
Amount $749,387.00
Phase II
Award Start Date 2021-09-07
Award End Date 2022-12-08
NewsGuard combines human intelligence with technology to enable
companies, institutions, and end-users to identify, track, and avoid
Abstract misinformation and disinformation online--and to find trustworthy sources
from which information is likely to be accurate.
The FOIA documents available for this award are heavily redacted. The disclosed text reflects that the
DOD used Misinformation Fingerprints to assess media outlets determined to be state-controlled in the following
283
General Questions, Small Business Innovation Research (SBIR) & Small Business Technology Transfer (STTR), U.S. Small Bus.
Admin. (last visited Jun. 29, 2024) available at https://www.sbir.gov/faqs/general-questions.
284
Eleanor Johnson, NSF 101: America’s Seed Fund, National Science Foundation (Nov. 2, 2023).
285
General Questions, Small Business Innovation Research (SBIR) & Small Business Technology Transfer (STTR), U.S. Small Bus.
Admin. (last visited Jun. 29, 2024) available at https://www.sbir.gov/faqs/general-questions.
286
Id.
287
NSIN Challenge – Countering COVID19 Disinformation, National Security Innovation Network (last visited Jul. 21, 2024)
available at https://nsin.mil/events/disinfo-challenge/.
55
ally markets: Australia, Canada, Germany, France, Italy, New Zealand, and the U.K.288
The Work Plan called for NewsGuard to conduct outreach to DOD’s partners with AI/machine learning or
social listening technology in order to integrate Misinformation Fingerprints into those partners’ programs.289 It is
not clear who these partners were.
The following facts evidence that the SBIR award was used to further develop the Misinformation
Fingerprints program, and that this is a government funding relationship, despite NewsGuard’s assertion otherwise:
1) The last of the award deliverables was to determine SBIR Phase III (commercialization) viability,
demonstrating growth from Phase II status.290
2) NewsGuard described its “work for the Pentagon’s Cyber Command;” doing “work” for the government
is not simply providing the government with access to existing proprietary data.291
3) Awardees of SBIR funds own, and have full right and title to, the data they develop under a SBIR
award.292
4) The award deliverable to develop a new field for example links to extract specific language from
outlets that contain the “myth narrative.”293
5) NewsGuard’s own framing of the award in its 2021 Social Impact Report indicated that the grant
would be used to “further develop” the Misinformation Fingerprints program).294
6) The following email in which NewsGuard staff asserts the award would be used to “enhance”
Misinformation Fingerprints.
295
The SBIR program has been a key tool for small businesses to access Federal resources to develop their
products. The program has distributed funds to small businesses for purposes that span from advancing research on
288
Phase II Work-Plan, FOIA of NewsGuard SBIR Award (Sep. 7, 2021) (on file with the Comm.).
289
Id.
290
Id.
291
Email from Gordon Crovitz, NewsGuard, Co-CEO, to Matt Taibbi, RacketNews, Investigative Journalist (Mar. 10, 2023).
292
Data Rights, Small Business Innovation Research (SBIR) & Small Business Technology Transfer (STTR), U.S. Small Bus. Admin.
(last visited Jun. 29, 2024) available at https://www.sbir.gov/faqs/data-rights.
293
Solicitation/Contract/Order for Commercial Items, FOIA of NewsGuard SBIR Award (Sep. 7, 2021) (on file with the Comm.).
294
295
Email from Matt Skibinski, General Manager, NewsGuard, to U.S. Dep’t of Defense (Aug. 3, 2021) (on file with the Comm.).
56
Alzheimer’s Disease to hypersonic turbulence modeling to monitoring for maritime search and rescue. The lasers
for Lasik eye surgery were developed using the SBIR program. However, over the last several years, the SBIR
program has also been leveraged to grow technology that surveils speech, ultimately used for the suppression and
censorship of that speech- with NewsGuard, this also applies to the press.
iv. Partnerships
NewsGuard’s partnership with State and the DOD was announced in August of 2020.296 By the end of
2021, NewsGuard had partnerships with more than 50 companies and organizations and more than 800 public
libraries.297 Their partners included search engines, social media platforms, global health agencies (including the
WHO), advertisers, news aggregators, media monitoring and reputation management companies, universities, the
U.S. government (including the Center for Disease Control and the Office of the Surgeon General), and schools.298
299
NewsGuard’s 2021 partnership with Microsoft made its Nutrition Label ratings technology available to
millions of internet users.300 It was a three-year agreement applying to many of its divisions, including Bing,
Microsoft News (MSN, a news aggregator), the Edge browser, Microsoft’s education and research departments,
and Microsoft’s Democracy Forward program.301 This partnership was promoted by the Disinfo Cloud Twitter
account.302
By this time, NewsGuard had partnered with three of the top five global advertising agencies and five
296
(Aug. 18, 2020).
297
298
Id.
299
Id. at 42.
300
Id. at 2.
301
Kate O’Sullivan, Microsoft’s General Manager of Digital Diplomacy, s on NewsGuard’s Global Advisory Board; Advisory Board,
NewsGuard (last visited Aug. 15, 2024).
302
Disinfo Cloud (@DisinfoCloud) reposting NewsGuard (@NewsGuardRating), Twitter (now X) (Jun. 14, 2021, 11:36 PM) avail-
able at https://x.com/newsguardrating/status/1404477925228007430?s=10&t=u8hEK0zejJhIrj8PBhzuqw.
57
of the top ten ad exchanges, which resulted in advertisers divesting from NewsGuard’s chosen “misinformation
economy,” and redirecting that ad spend to outlets NewsGuard deemed trustworthy.303 It had signed partnerships
with three of the top six advertising holding companies, representing more than 100 agencies.304 NewsGuard’s
reliability ratings influenced thousands of ad buys.305
NewsGuard secured partnerships that spread the use of its ratings beyond programmatic display advertising
on news and information sites to also cover advertising on YouTube, extending its reach to social media
monetization.306 By partnering with Zefr, a “brand safety” technology company used by YouTube, NewsGuard
was able to influence brands to target their YouTube ads to sources on the platform NewsGuard deemed credible.307
During the pandemic, the WHO enlisted NewsGuard for its input, including regular reports, on which
COVID-19 narratives it determined to be misinformation were prevalent online.308 This included information
about the virus itself, the vaccines, and treatments. The WHO then contacted social media companies and search
engines asking them to remove this content.309
By the end of 2022, NewsGuard had secured significantly more partnerships.310 Through its relationship
with IPG Mediabrands, NewsGuard expanded its ratings system to 140 cable and streaming TV shows and
networks.311 This partnership was also promoted by Disinfo Cloud’s Twitter account, along with an article
discussing how IPG Mediabrands provided its clients with discounted rates for using NewsGuard’s data and
recommended only networks scoring 85 or higher on NewsGuard’s scale.312 Mediabrands entities UM, Initiative,
and MAGNA worked together on the deal.313 Other partnerships were secured with ad-tech platforms such as
Media.net, Unruly, and Magnite. NewsGuard launched its Responsible News Private Marketplace with Pubmatic,
which provided brands with the ability to buy inventory for more than 3,000 news sites they deemed credible.314
By the end of 2023, NewsGuard’s had secured significantly more clients and partnerships:
303
304
Id. at 20.
305
Id.
306
Id. at 26.
307
Id.
308
Id. at 7.
309
Id. at 4.
310
311
Id. at 10.
312
Disinfo Cloud (@DisinfoCloud) reposting NewsGuard (@NewsGuardRating), Twitter (now X) (Dec. 2, 2021, 10:02 AM) avail-
able at https://x.com/newsguardrating/status/1466422604810670092?s=10&t=u8hEK0zejJhIrj8PBhzuqw; IPG Mediabrands Strikes
Exclusive Deal with NewsGuard to Go Beyond Website Ratings to Rate Individual Cable and Broadcast TV News Shows, NewsGuard
(Dec. 2, 2021).
313
IPG Mediabrands Strikes Exclusive Deal with NewsGuard to Go Beyond Website Ratings to Rate Individual Cable and Broadcast
TV News Shows, NewsGuard (Dec. 2, 2021).
314
58
315
NewsGuard now leverages relationships with Meltwater, PeakMetrics, and Pulsar to “help brands
safeguard their reputations” online.316 Its data supports media monitoring, public relations, social listening, and
reputation management companies advise clients about the “quality and trustworthiness” of websites and social
media channels the companies appear on.317
While more transparent than the GDI, NewsGuard’s reach is far more significant in impact and scope.
Running afoul of its rating system or refusing to implement the reforms NewsGuard demands in order to raise
scores has marked effects on a businesses’ reputation, circulation, revenue, and overall operations. This is especially
harmful to small businesses that cannot survive such impediments to growth. That advertising organizations
were encouraging their members to use NewsGuard (and GDI) may have antitrust implications, as demonstrated
by the House Judiciary Committee.318 Those agreements among the private sector to drive revenue away from
disfavored press, combined with the backing of the Federal government, is an unacceptable interference in the
free marketplace of ideas. The Federal government cannot, in any way, be involved in this space.
315
Social Impact Report 2023, NewsGuard, 21-23 (Jan. 2024).
316
Id. at 17.
317
Id.
318
H. Comm. on the Judiciary, GARM’s Harm: How the World’s Biggest Brands Seek to Control Online Speech (Jul. 10, 2024).
59
V. Recommendations
1. Investigation into Other GEC Awards
The Committee has received only a fraction of the documents it requested and subpoenaed from the GEC
and therefore has not performed a thorough review of suspect grants. The GEC awards discussed in this report are
not the only ones at issue. Some award recipients, such as the Alliance for Securing Democracy and the Atlantic
Council and its DFRLab, have been examined in depth by other Congressional committees and investigative
journalists; most have not been thoroughly reviewed. There is much more to learn and transparency is in short
supply. Further inspection should be carried out into GEC awardees, beginning with those with known domestic
impacts.
2. Audit of GEC Award Records and State Records Writ Large

Throughout this investigation it became clear that the GEC has several issues in its recordkeeping and
that there are not sufficient audit procedures in place to efficiently track its use of taxpayer dollars. The following
issues were present in the subaward records alone: (1) categories were provided for several recipients rather
than specific organizations or individuals, such as $240,136 for “Radio Programmes” and $42,600 for “On-Air
Discussion;” (2) in six instances, subawardees were just the first names of individuals; (3) in one instance the
field denoted “Report mentions subpartners; unable to find details;” and (4) it appears at least one subaward was
omitted, as only one Park Advisors subaward to NewsGuard is shown for $50,000, and there are two known Park
Advisors subawards to NewsGuard, the other for $25,000.
Despite State’s claims that it conducts rigorous oversight of subawardees, the GEC’s records indicate that
there are numerous subawardees for which State has little or no information. Fears of a more widespread issue are
exacerbated by the July 2024 Special Inspector General report which found that USAID could not demonstrate
compliance with its award requirements for at least $239 million in taxpayer dollars distributed in Afghanistan,
some of which may have been funneled to the Taliban.319 Though an independent Federal government agency,
USAID receives overall policy guidance from the Secretary of State.
In its analysis of the GEC’s productions, the Committee found ten publicly available GEC direct awards
that were not disclosed. After nearly one year of discussion with State, in which it failed to offer an explanation
or perform internal due diligence to resolve the discrepancy, State realized that these ten awards were mistakes;
each was improperly logged as a GEC award by various embassies. Suffice it to say that it should not fall to the
Committee on Small Business to catch errors in the Department of State’s record keeping. If, as State told the
Committee on December 19, 2023, these award numbers are “tracked and used for reporting,” there are not audits
or procedures in place sufficient to catch these errors.320 If this Committee’s narrow review of recent GEC awards
alone found ten errors, it is reasonable to conclude similar errors exist department-wide.
There needs to be more efficient auditing of taxpayer dollars that are given away. A complete review of
State’s awards, at all levels, is needed to ensure taxpayer dollars are being used for their proper and intended
purpose, and granted to the intended awardee, with proper documentation.
319
Special Inspector Gen. for Afghanistan Reconstruction, SIGAR 24-29-FA, USAID’s Afghanistan Conflict Mitigation Assis-
tance for Civilians Program: Audit of Costs Incurred by Blumont Global Development, Inc. (Jun. 2024).
320
Email from Molly Claflin, Special Counsel, U.S. Dep’t of State, to Stephanie Chambless, Counsel, H. Comm. on Small Bus (Dec.
19, 2023).
60
3. SBIR Program Reforms
The SBIR program is America’s seed fund and should not be used to grow companies (like Meedan) that
surveil the speech of American citizens or those (like NewsGuard) that seek to demonetize domestic press. The
Committee remains concerned about this direction of the SBIR program, and whether the SBA, as the agency in
charge of directing program policy, has had anything to do with this shift in award theme. The SBIR program is
due for reauthorization in FY 2025. Guardrails should be put in place to prevent these types of companies from
receiving taxpayer dollars and to facilitate effective oversight and greater transparency into the SBIR program.
Further, the May 2024 report on the origins and development of the GEC states that in 2016 the SBA was
part of the GEC’s Steering Committee.321 It is not clear whether this relationship still exists. Between the SBA’s
role administering the SBIR program and the GEC’s disinformation-oriented awards, the link between the SBA
and the GEC is suspect. The Committee seeks greater transparency on the relationship between the SBA and
GEC.
4. Oversight of the NED

The NED’s use of taxpayer dollars is suspect and should undergo Congressional oversight and review
as permitted by the National Endowment for Democracy Act. The NED was appropriated approximately $300
million from the government in FY 2022.322 This is not a small amount of money, and when a supposedly bipartisan
organization with international restrictions has staff using their NED email to discuss the legitimacy of domestic
press outlets, combined with funding of organizations like the GDI, it is indicative of rot.
5. Legislation
No Federal funds should be used to grow companies whose operations are designed to demonetize and
interfere with the domestic press. Though the government is no longer in a relationship with NewsGuard or the
GDI, the damage has already been done—they have already received the backing of the Federal government in
hosting their products on the GEC’s Testbed and recommending them to its partners, using their services, and
helping to grow their products. The only real change the Federal government could be involved in moving forward
would be to restrict future Federal funding to them and companies with similar operations. Congress should also
explore the impacts of cutting off Federal funding to entities who use these products.
Further, the DOD should not be outsourcing their perception of fact to speech-police organizations subject
to partisan bias. For this reason, Chairman Williams introduced an amendment to the FY 2025 NDAA to block
such funding, which passed the House with a unanimous Republican vote, despite every single Democrat voting
‘no.’323 Members, Committees, and their staff must remain diligent about government programs and contracts
which threaten to undermine the First Amendment rights of Americans and domestic press.
321
322
Financial Report, National Endowment for Democracy (Sep. 30, 2022).
323
Amendment to Rules Committee Print 118-36, Offered by Roger Williams (Jun. 6, 2024) available at https://amendments-rules.
house.gov/amendments/WILLTX_062_xml240607115518651.pdf.
61
VI. Conclusion
The collaboration between the public and private sector to moderate certain content impacts not only
the First Amendment rights of individuals, but the ability of businesses to compete in a free marketplace. This
interference is especially harmful to small businesses, which need to be able to freely utilize all available
avenues for growth. The ability to leverage various internet platforms is essential for competition in the modern
marketplace. This manner of government interference results not only in economic harms, but also brushes up
against foundational, Constitutional principles.
In the dissent for Murthy v. Missouri, Justice Alito describes the difference in vulnerability to government
pressure between social media platforms and news sources. He explains that social media companies are critically
dependent on the protection provided by §230 of the Communications Decency Act of 1996 and are susceptible to
antitrust lawsuits, and are therefore far more vulnerable to government pressure.324 He argues that “[i]f a President
dislikes a particular newspaper, he (fortunately) lacks the ability to put the paper out of business.”325 This is meant
to be an assurance, but with the Federal government supporting and promoting companies whose products aim to
demonetize news outlets and disrupt their operations because of their lawful speech, this claim is not necessarily
true.
Taxpayer dollars and other Federal support should not be given to organizations who police lawful
speech and punish domestic press noncompliance with their standards and points of view. Press rating systems
are inherently flawed as they are subject to the partisan lens of the assessor. News and information businesses
must be able to compete online in a free and fair marketplace without having to conform to narratives pushed by
government-backed ratings organizations for fear of losing advertising revenue. It is extremely concerning that
such organizations, in addition to hundreds of products that surveil internet speech and decide what is truth, were
given government platforms and Federal funding to grow and promoted directly to the social media companies
in charge of moderating online speech. It is similarly concerning that this technology was promoted to foreign
governments with significant internet speech regulations and to Federal agencies without international restrictions
and whose coordination with social media is suspect at best.
Inaccurately painting speech as false has become a dangerous recurring theme. MDM labels are being
weaponized against dissident narratives. Speech is being framed as true or false often on a partisan, rather than
factual, basis. While actual false information certainly can inflict harm, the solution should not be to censor it, but
to combat it with correct information. Open dialogue is necessary to resolve complex issues. The elimination of
dissenting voices is not the solution. As George Washington famously said:
“For if men are to be precluded from offering their sentiments on a matter, which may involve the most serious
and alarming consequences […] the freedom of speech may be taken away- and, dumb and silent we may be
led, like sheep to the slaughter.”326
324
Murthy, et al. v. Missouri, et al., 603 U.S. 39 (2024) (Alito, J., dissenting).
325
Id.
326
George Washington, The Newburgh Address (Mar. 15, 1783).
62
VII. Appendix
Before FY 2019 (pre- 9/30/2018)
• 4/2007: Counterterrorism Communication Center (CTCC) (first precursor to the GEC) established
• 2008: CTCC rebranded as the Global Strategic Engagement Center (GSEC)
• 1/20/2009: Barack Obama sworn into office
• 9/2010: Center for Strategic Counterterrorism Communication (CSCC) replaces the GSEC
• 9/9/2011: EO 13584 officially establishes the CSCC with a mission and mandate327
• 1/2016: Obama White House announces creation of the GEC to replace the CSCC
• 3/14/2016: EO 13721 establishes the GEC and its multi-agency Steering Committee
• 12/23/2016: FY 2017 NDAA signed into law, expanding the GEC’s mission by giving it the authority to
address other foreign propaganda and disinformation operations
• 1/20/2017: Donald Trump sworn into office
• 2/23/2018: Funds transferred from the DOD to the GEC for initiatives to counter propaganda and
disinformation from foreign nations; included creation of the IAF
• 8/13/2018: FY 2019 NDAA signed into law, further expanding GEC’s scope of work, endowing it with a
mandate, as reflected in its current mission statement328
• 9/25/2018: Start date of Park Advisors GEC award
FY 2019 (10/1/2018 – 9/30/2019)
• 11/26/2018: CISA established within DHS
FY 2020 (10/1/2019 – 9/30/2020)
• 7/1/2020: Start date of first NED GDI award
FY 2021 (10/1/2020 – 9/30/2021)
• 11/2/2020: Contract signed for GEC/DOD split GDI award

• 1/20/2021: Joe Biden sworn into office
• 7/1/2021: Meedan SBIR Award Start Date
• 7/31/2021: Applications close for U.S.-Paris Tech Challenge
• 9/7/2021: Newsguard SBIR award Start Date
FY 2022 (10/1/2021 – 9/30/2022)
• 12/31/2021: Contract signed for GEC-standalone GDI award

• 12/31/2021: End date of Park Advisors GEC award
• 4/14/2022: Elon Musk buys Twitter
• 9/2022: State’s OIG found GEC not properly overseeing contractors to ensure they do not perform
“inherently governmental functions”329
FY 2023 (10/1/2022 – 9/30/2023)
• 10/1/2022: Start date of final NED GDI award
327
Exec. Order No. 13584, Developing an Integrated Strategic Counterterrorism Communications Initiative, 76 Fed. Reg. 56945 (Sep. 9,
2011).
328
John S. McCain National Defense Authorization Act for Fiscal Year 2019, Pub. L. No. 115-232, § 1284, 132 Stat. 1636 (2018) available
at https://www.congress.gov/bill/115th-congress/house-bill/5515/text.
329
U.S. Dept. of State, Office of the Inspector Gen., ISP-i-22-15, Inspection of the Global Engagement Center (Sep. 2022).
63
• 10/21/2022: GDI releases list of 10 riskiest/10 least risky American media outlets
• 12/3/2022: First installment of the Twitter Files released
• 12/16/2022: GDI publishes report of riskiest American media outlets
• 2/17/2023: NED terminates all GDI awards
• 3/2/2023: Date of 2023.02.14 GEC-GDI-BLACKLIST.docx
• 3/2/2023: Twitter Files on the GEC released (#17)
• 3/9/2023: First Weaponization subcommittee hearing with Twitter Files authors Matt Taibbi and Michael
Shellenberger as witnesses
• 3/10/2023: Newsguard emails Matt Taibbi to dispute the “government funded” categorization
• 6/7/2023: HCSB Letter 1 to GEC
• 7/24/2023: Second HCSB Letter 2 to GEC
FY 2024 (10/1/2023 – present)
• 11/30/2023: Second Weaponization hearing with Matt Taibbi & Michael Shellenberger
• 12/3/2023: State Production 1 listing direct awardees FY 2019 – FY 2023
• 12/6/2023: State of Texas/The Federalist/The Daily Wire file suit against State/GEC
• 12/19/2023: In-Camera review of Production 1
• 2/14/2024: GEC Letter to HCSB re: disclosure
• 3/21/2024: HFAC Subcommittee hearing on oversight of the GEC
• 3/28/2024: State Production 2 listing subawardees FY 2018 – FY 2023
• 4/9/2024: Briefing with GEC grant officers and grant officer representatives
• 5/2/2024: In-Camera review of Production 2
• 5/20/2024: GEC responds to Rep. Beth Van Duyne HFAC QFRs
• 6/13/2024: HCSB issues subpoena to State
• 6/14/2024: GEC responds to questions about Federal Acquisition Identification Numbers
• 6/26/2024: HCSB GEC/Censorship hearing
• 6/27/2024: GEC produces 2023.02.14 GEC-GDI-BLACKLIST.docx
• 7/24/2024: Judiciary report on GARM and possible antitrust violations
• 8/8/2024: GEC produces information on DT Institute
• 8/27/2024: GEC produces information on GLOBSEC and East-West Center
64
Organization Abbreviation
Atlantic Council’s Digital Research Forensics Lab DFRLab
Application Programming Interface API
Censorship-Industrial Complex CIC
Center for Disease Control CDC
Central Intelligence Agency CIA
Congressional Research Service CRS
Counterterrorism Communication Center CTCC
Counter propaganda and disinformation CPD
Cybersecurity and Infrastructure Security Agency CISA
Center for Strategic Counterterrorism Communications CSCC
Department of Defense DOD
Department of Homeland Security DHS
Department of Justice DOJ
Department of State State
East-West Center EWC
Executive Order EO
Federal Award Identification Number FAIN
Federal Bureau of Investigation FBI
Fiscal Year FY
Freedom of Information Act FOIA
Global Alliance for Responsible Media GARM
Global Disinformation Index GDI
Global Disinformation Index’s Dynamic Exclusion List DEL
Global Engagement Center GEC
Global Strategic Engagement Center GSEC
House Committee on the Judiciary Judiciary or HCJ
House Committee on the Judiciary, Subcommittee on the
Weaponization
Weaponization of the Federal Government
House Committee on Small Business Committee or HCSB
House Foreign Affairs Committee HFAC
Information Access Fund IAF
Institute for War & Peace Reporting IWPR
Letter from HCSB to GEC on June 7, 2023 Letter 1
Letter from HCSB to GEC on July 24, 2023 Letter 2
Letter from HCSB to GEC on January 28, 2024 Letter 3
Letter from HCSB to GEC on February 16, 2024 Letter 4
Letter from HCSB to GEC on May 16, 2024 Letter 5
Misinformation, Disinformation, and Malinformation MDM
National Defense Authorization Act NDAA
National Endowment for Democracy NED
65
National Science Foundation NSF
New York Times NYT
Nongovernmental organization NGO
North Atlantic Treaty Organization NATO
Office of the Director of National Intelligence ODNI
Poynter’s International Fact-Checking Network IFCN
Production from the GEC to HCSB on December 3, 2023 Production 1
Production from the GEC to HCSB on March 28, 2024 Production 2
Research & Development R&D
Small Business Administration SBA
Small Business Innovation & Research program SBIR
Software-as-a-Service SaaS
Technology Engagement Team TET
Terms of Service TOS
Twitter (now X) Twitter330
Questions for the Record QFR
U.K. Foreign Commonwealth, and Development Office U.K. FDCO
U.S. Agency for International Development USAID
World Health Organization WHO
330
As the majority of the events in this report took place prior to Twitter’s rebranding as X, that platform is referred to as Twitter
throughout this report.
66

House Small Business Committee Interim Report 2024

Uploaded by

Copyright:

Available Formats

House Small Business Committee Interim Report 2024

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

House Small Business Committee Interim Report 2024

Uploaded by

Copyright:

Available Formats

Small Business:

Instruments and Casualties of the

The House Committee on Small Business

2. The Technology Engagement Team & Private Sector Engagement

Objective Award Purpose Subawards Subaward Amount

i. Testbed Methodology & Disinfo Cloud

a. 365 hosted technologies;

ii. Tech Demo Series

Date of Company Product Product Description

Date of Company Product Product Description

iii. International Tech Challenges

B. Global Disinformation Index

i. About the GDI

a. Disinformation Index Ltd.: a U.K. private company limited by guarantee;

b. Dynamic Exclusion List

NED1 2020-10474 (2020-1116) AN Foundation 7/1/2020 2/17/2023 $230,000

1) Albany Associates International (office in Bethesda, MD);

a. Reliability Ratings & Nutrition Labels

1) 22 points: does not repeatedly publish false or egregiously misleading content.

Accompanying each “Fingerprint” is the following data:

1) Example language used to advance the narrative;

c. NewsGuard for Advertising

d. Social Impact Reports

iii. U.S. Government Funding

a. GEC/DOD Split Award

c. DOD SBIR Award

Recipient NewsGuard Technologies, Inc.

2. Audit of GEC Award Records and State Records Writ Large

4. Oversight of the NED

Before FY 2019 (pre- 9/30/2018)

FY 2019 (10/1/2018 – 9/30/2019)

• 11/26/2018: CISA established within DHS

FY 2020 (10/1/2019 – 9/30/2020)

• 7/1/2020: Start date of first NED GDI award

FY 2021 (10/1/2020 – 9/30/2021)

• 11/2/2020: Contract signed for GEC/DOD split GDI award

FY 2022 (10/1/2021 – 9/30/2022)

• 12/31/2021: Contract signed for GEC-standalone GDI award

FY 2023 (10/1/2022 – 9/30/2023)

• 10/1/2022: Start date of final NED GDI award

FY 2024 (10/1/2023 – present)

You might also like