N D R 1.

OPIC: E HICAL IMPLICA IONS OF SING AI IN ED CA ION

SPONSORS: People s Republic of C and Republic of I

SIGNA ORIES: Islamic Republic of I , The Republic of S A , Islamic

Republic of P , Republic of A , Republic of K , The Kingdom of
S , United Republic of , The Commonwealth of A , A
E , Republic of , The Arab Republic of E , The Republic of I ,
The Republic of P , The Kingdom of M , the Arab Republic of E , the
Kingdom Of S , The Republic of P , the Federal Republic of N , the
Kingdom of N , Republic of , the Republic of P , The People s
Republic of B , the F Republic, Federative Republic of B , The
I Republic, State of J , Dominion of C , Socialist Republic of ,
Federal Republic of G , United M States,

The U i ed Na i Ed ca i a , Scie ific a d C a O ga i a i

G ided b he Uni ed Na ion Cha e and i inci le of omo ing ocial og e , ali of
life, and ba ic h man igh ;

Recogni ing UNESCO manda e of Ed ca ion fo Al, Ca aci B ilding and S ainable
De elo men , E hical S anda d in Science and Technolog ;

Em ha i ing he need o achie e ainable de elo men goal ;

Ackno ledging he diffe en c al ole ha each co n ha ,

Recall he a ac ion made b UNESCO ch a b no limi ed o Recommenda ion on he

E hic of AI, UNESCO Global AI E hic and Go e nance Ob e a o , and Readine
A e men Me hodolog (RAM)

AR ICLE 1: ADDRESSING HE E HICAL ISS ES REGARDING HE SAGE OF AI

IN ED CA ION

1. Promotes member states to utilize and implement the UNESCO s Recommendation of

the Ethics of Artificial Intelligence as a guidance to implement their comprehensive
framework of laws within the country;
a. Defines "ethical usage of AI to consist of interaction with AI systems that are not
harmful but contribute to peace, human dignity, sustainability, and security for
students, educators, and administrators, hence should strive to maintain data
privacy and security, identify and redress such biases to ensure fair and inclusive
education and, ensure that AI technologies are inclusive and accessible to all;
2. Recommends the implementation of C L AI M
(CLAIM) to ensure the adherence of AI companies, especially those operated by foreign
investors, to the applicable international guidelines such as but not limited to UN Data
Privacy, Ethics Guidance and national standards and regulations; This licensing is aimed
at ensuring the safety and reliability of AI system provides, so they can provide fast,
accurate, and transparent services in the educational sector, through the following
mechanisms such as:
a. Calls upon member states to work in close coordination with respective
ministries, such as ministry of education and MOI, to establish a regulatory
requirement and conduct system testing;

i. The requirement will include several points such as a commitment by operators to ensure
responsible data security in accordance with applicable laws;
 ii. AI companies must provide broad explanation of electronic system s operation;

b. Acknowledges the principle of data nationalism as an effort by states to ensure

control over data for a range of normative and security-based reasons, and data
sovereignty that such personal data collected in geographic location must be
subject to the laws on that location;

3; Supports the development of ethical guidelines and standards for the use of AI in education
taking into account diverse cultural and societal contexts by the following mechanisms such as
through BERR (Building AI ethics curricula, Ethics research, Researchers training, Implement
EIA)I:

a. Building AI Ethics Curricula through the collaboration with UNICEF to promote the
integration of AI ethics education and research by integrating AI ethics into national
education programmes at all levels; This includes creating AI ethics curricula that cover
the technical aspects of AI and that are sensitive to local cultures and values, ensuring
that educational materials are available in local and indigenous languages, and ensuring
accessible for students such as through websites;
b. Support AI thics research by creating incentives for both the public and private sectors
to invest in AI ethics research through;
i. Funding Allocation: Allocate multilateral funds to support AI ethics research
projects, both within national research institutions and in collaboration with
international partner;
ii. Grants and Awards: Establish grants and awards for outstanding achievements in
AI ethics research to recognize and reward individuals and organizations that
make significant contributions to the field;
iii. Public-Private Partnerships (PPP) and Corporate Social Responsibility (CSR):
Facilitate PPP and CSR that focus on AI ethics research, leveraging the strengths
of both sectors to drive innovation and ethical standards in AI; Through this,
projects will be finished faster;
 iv. Capacity Building: Invest in capacity building programs to train researchers and
professionals in AI ethics, ensuring a skilled workforce that can address the
ethical challenges of AI;
4; Encourages the program for esearchers training to ensure that AI researchers receive training
in research ethics as part of their education and professional development; This training should
equip researchers with the knowledge and skills necessary to incorporate ethical considerations
into every stage of their work, from the design of AI systems to the publication of their research
findings;
5; Encourages member states to conduct E I A (EIA) which considers the
entire process of designing, developing and deploying an AI system allowing for assessment of
the risks before and after the system is released to the public; EIA is very important in ensuring
the ethical design and use of AI; AI systems and tools have often been released to users without
clear and transparent analysis of the potential risks, risk mitigation, and addressing unintended
consequences; EIA aims to fix that problem by furnishing both private and public sectors with
the set of questions to ask in order to ensure that the AI systems they are purchasing or
developing are aligned with the ethical standards set out in UNESCO s Recommendation on
Ethics of AI; This includes several activities such as but not limited to regulatory sandbox and
risk assessments;

AR ICLE 2: INCL SI I , BIASED-FREE DA ABASE, AND COLLEC ION

S S EM

1. Promotes an inclusive, foreign-biased-free database and collection system through project

MANGO (M A B A GIGO), through the following mechanisms
such as:
a. Mitigate algorithmic bias by the following mechanisms such as:
i. Provide more diverse data to training datasets, ensure data represents the
whole demographic, to increase representation and reduce bias that can be
achieved through data augmentation as the process of artificially
generating new data from existing data, primarily to train new machine
learning (ML) models;
 ii. Incorporating the method of oversampling the underrepresented
group/data/ sources to prevent narrow databases that AIs may use;
Research shows that oversampling underrepresented groups can help
create more diverse output; Thus to solve this problem can oversample
data from local language cultures as well as local people;
b. Recognizes that the quality of output in this case of AI algorithms is only as good
as the quality of input, it is essential to maintain good, healthy training datasets to
avoid the pitfalls of "G I , G O " (GIGO) via designing
algorithms that consider different types of bias and aim to minimize their impact
on the system s outputs;
c. Conducts yearly periodic reviews, evaluations, and verifications of the
algorithms mechanism and principles, models, data, and application results by
national experts and to complement EIA report made by UNESCO;
2. Encourages for member states to develop national action plans concerning information
security and reliability, under each countries respective N AI R &
A (NAR O) which aims to ensure transparency and
security of the utilization of AI technologies in the educational sector with the following
mechanisms such as:
a. Transparency and accountability
i. Encourage countries to send a compendium of good practices which
contains progress, challenges and success in the utilization of AI in
education which will be reviewed by the working bodies under UNESCO
which will serve as a guideline for future allocation resources or aids;
These working bodies consist of technical experts from different nations
but within the supervision of the United Nations;
ii. The annual report may contain, such as, but not limited to; source of data,
goals of using AI, progress of AI implementation, hindrance of AI usage,
user experience;
b. Preventing bias
 i. Initiate collaboration with local government and local organizations to further
consult related to stereotypes of languages or related to local language to prevent
any discrimination towards certain language;
ii. Ensure the diversity of data used to be incorporated into the machine-learning
models through monthly surveys to prevent any bias that is raised due to
incomplete data, and implement tools which can assess the percentage of bias
from the result;
iii. Enable human oversight to monitor and intervene the scoring systems and
students assessments to evaluate the bias or errors that comes from AI tools

AR ICLE 3: NDERS ANDING HE ISS E REGARDING DA A PRI AC AND

SEC RI IN AI-ED CA IONAL PLA FORM

1. Recommends member states to implement security verification to AI models, aims to

prevent data leakage, security breach, and other foul misuse of AI, through the following
methods such as but not limited to:
a. Up-to-date firewalls;
b. Access control such as but not limited to through multi-factor authentication;
c. Improve IT industry self-regulatory;
d. Data encryption and anonymization;
e. End-point encryption;
f. Built-in data protection;
2. Encourages the collaboration with the United Nations Sustainable Development Group
(UNDG) to endorse UNDG s DATA PRIVACY, ETHICS AND PROTECTION
GUIDANCE NOTE as a universal minimum data privacy and security standards that
uphold lawfulness, fairness, transparency, accuracy, and accountability, through the
following specifications such as:
a. Upon which each member state within UNESCO is encouraged to use as a basis
to implement its own laws that contains explicit regulations to accommodate each
country s sovereignty;
 b. Principles such as what constitutes consent from a data subject and recognizes
privacy rights for data subjects including the right to data portability, the right to
object, and the right not to be subjected to automatic decisions;
3. Recommends this following programs to address the risks and harms of large amounts of
data used by AI;
a. Ensure data security and protection through DA E: D -
A E ;
i. Data De-identification: De-identify personal data where appropriate,
de-identification means a process of using all reasonable means to convert
personal data into anonymous data, such as that it cannot be traced back or
linked to an individual(s) or group(s) of individuals; using methods such
as aggregation, pseudonymization, or masking, to minimize potential risks
to privacy;
1. Data aggregation is the process of combining individual data sets
in such a way that they can no longer be traced back to or linked to
any specific individual; In other words, the International
Organization for Migration (IOM) defines “aggregate data in its
Data Protection Manual as summary statistics compiled from
personal data but grouped to prevent identifying individual cases;
2. Masking means a de-identification technique whereby the original
personal information collected from social media, such as
comments, photos and videos, is altered to such an extent that it
cannot be traced back or linked to an individual(s) or group(s) of
individuals;
3. Pseudonymization means modifying personal data by removing or
substituting all direct identifiers (e;g; in many instances name,
address, date of birth, etc;) with other unique identifiers (e;g; in
many instances hashing algorithms, ID numbers, etc;)
b. Encryption: Encrypt personal and sensitive data when transferred to or from any
network-connected servers
4. Recommends NAR O data privacy mechanisms, through the following specifications:
 a. Allocate more resources to improve data security infrastructures, such as, but not
limited to data security, firewall, encryption and strong network security;
b. Encourage the implementation of regular auditing of technological companies
which will be done by special units under relevant authorities of each country to
minimize risks of data leakage and privacy infringement;

AR ICLE 4: NO ING HE ISS E OF EQ I ABLE DIS RIB ION AND DIGI AL

DI IDE

1. Encourages development of GRAPE: G R AI P E , for

digital training and resources, through the following mechanisms such as:
a. Encourages the creation of teacher training programs specifically designed to
educate teachers in rural, non-prestigious schools to equip them with knowledge
and skills, and digital literacy on the ethical guidelines and application of AI in
education, emphasizing on the principle that AI should support human actions and
not supplant them;
b. Promotes the development of AI through Public-Private Partnerships (PPP) over
the next two years in which the government invests in private companies (foreign
and domestic) to develop AI; After the AI is developed, use EIA to assess the
potential risks of the AI after which the local government can take over the
deployment and regulation of the AI to ensure the developed AI is tailored to the
specific needs and context of the national interest and local economy;
2. Promotes Empowerment of less technologically advanced countries by A
C - H (PEACH) through the following mechanisms such as:
a. Endorse the usage of capacity building between technologically advanced and less
technologically advanced countries by:
i. Implement UNESCO s Readiness Assessment Methodology (RAM) to
evaluate a country's preparedness for implementing AI; RAM includes a
range of quantitative and qualitative questions designed to gather
information about various dimensions of a country s AI ecosystem,
 including legal and regulatory, social and cultural, economic, scientific
and educational, and technological and infrastructural aspects;
ii. The purpose is to identify the strengths and weaknesses of a country's AI
ecosystem, enabling targeted improvements and facilitating the
development of a robust national AI strategy; This assessment will help
bridge the digital divide by providing less technologically advanced
countries with a clear roadmap to enhance their AI capabilities and foster
sustainable development; Furthermore, it allows UNESCO to tailored its
recommendations based on the needs of each country;
iii. Then after completing RAM, the country will get a report about the status
of AI readiness in the country, summarizing where the country stands on
each dimension, detailing ongoing activities, summing up the state of the
art, and providing concrete policy recommendations on how to address
governance gaps; This report helps identify in detail what institutional
changes are required to elaborate or bolster national AI strategy;
iv. Based on the findings from the RAM report, technologically advanced
countries with extensive AI policies can provide expertise, ethical
AI-policy insights, and training programs tailored to less technologically
advanced countries; These efforts will build local capacity, empower less
developed countries to utilize AI effectively, and promote sustainable
development through equitable access to advanced technologies;
3. Recommends the implementation of A I R D
O S (AIRDROP) for technologically advanced countries to
engage in bilateral and / or multilateral cooperation with less-technologically advanced
countries to assist the integration of AI, these includes several points such as:
a. Transfer of knowledge through share best practices, provide guidance in the
formulation and establishment of AI legislations, organize workshops and
seminars for the dissemination of knowledge;
b. Technical assistance such as but not limited to dispatching AI experts and
engineers;
c. Transfer of technology such as but not limited to development of internet;
4. Recommends to expand UNESCO teachers training course concerning AI ethics in
education as a part of a multi-stakeholder literacy efforts aimed at making the future AI
technology responsible and promote the best practices, through the following
specifications;
a. The course can be delivered through Focus Group Discussions (FGD) and
teaching simulations on AI in classrooms;
b. This course helps teachers create inclusive learning environments with the
students, such as, how to communicate ethical dilemmas related to the use of AI;
c. This training can be provided in cooperation with educational institutions, such as
universities and schools;
5. Recommends B D D (BRIDGE) Program to ensure all students
have equal access to advanced technologies and digital resources for learning, this
includes infrastructure development to lay the stable groundwork required to implement
technology in education, and enhanced internet access, through the following
specifications;
a. Calls upon the collaboration among local government, educational personnel, and
local communities to work together to foster a supportive ecosystem in learning
progress;
b. This program can be supported by collaboration with developed countries and
private sectors through PPP for funding sustainability;
c. Encourages the collaboration with UNDP (United Nations Development
Programme) whose commitment aims to reduce inequalities between groups and
people;
6. Recommends a legal framework called G &N G
H R D AI E (G ANG HO F )
where it focuses to provide people the basic knowledge of AI technologies and train them
to better utilize AI tools in their daily lives, particularly in the education sector which will
enact 4 phases to integrate countries which are on different levels of readiness to
incorporate AI into their educational system;
a. PHASE 1: C ; This phase will primarily focus on preparing educators to
teach AI-integrated lessons at school levels through training programmes with the
following mechanisms such as:
i. Encourage the provision of training programmes for teachers in schools to
introduce the utilization of AI tools to assist them on daily basis through, but not
limited to; private workshops and public conferences;
ii. Recommends the integration of AI education in school curriculums by
reviewing the readiness and resources of schools in local regions through
school-to-school surveys through these following mechanisms;
iii. Encourage each representative from schools in the local region to visit an
AI-incorporated schools and can learn to further implement AI into their own
classrooms;
iv. Encourage reviewers to come to each school to check current resources and
report materials that must be improved;
b. PHASE 2: I AI; This phase will prioritize on the application of AI
technologies as a helping tool to assist educators to ensure a smoother transition of a
digitized education, especially in less developing areas with the following mechanisms:
i. Encourage the integration of AI as part of mandatory tools in curriculum when
learning certain subjects, but not limited to; visual learning, auditory learning,
verbal learning, imaginative learning;
ii. This phase is intended to raise awareness to attract prospective students
interested in AI within the youth about the usage of AI which will be further
expand the next phase;
c. PHASE 3: E AI; This phase will mainly focus on promoting and socializing
the usage of AI technologies in the country through the educational sector to prepare
people to enter the AI market, through mechanisms such as:
i. Encourage the integration of AI-related courses and programs across
disciplines, covering fundamentals, applications, and ethics of AI, and develop
specialized tracks or concentrations in AI for interested students;
ii. Suggest the foster partnerships with industry leaders and AI practitioners to
offer guest lectures, internships, and collaborative research opportunities to
enhance practical learning experiences and keep curriculum relevant to industry
needs;
iii. Provide opportunities to enhance human resources by encouraging countries
who are already AI-advanced to open online courses or scholarship opportunities
for developing countries who are lacking assistance in developing their AI
knowledge.