See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/350734729

Combining Blockchain and Artificial Intelligence - Literature Review and State

of the Art

Conference Paper · November 2020

CITATIONS READS

0 37

1 author:

Erik Karger
University of Duisburg-Essen
3 PUBLICATIONS   0 CITATIONS   

SEE PROFILE

All content following this page was uploaded by Erik Karger on 08 April 2021.

The user has requested enhancement of the downloaded file.

 Combining Blockchain and Artificial Intelligence

Combining Blockchain and Artificial

Intelligence – Literature Review and State of
the Art
Completed Research Paper

Erik Karger
University Duisburg-Essen
Universitätsstraße 9, 45141 Essen
[email protected]

Abstract
Artificial intelligence and blockchain are among the most popular technologies. Combine the
two technologies also harbors manifold potentials. For instance, blockchain can help address
specific AI-related difficulties such as the black box problem. Vice versa, AI offers opportunities
to improve the blockchain’s mining process, or smart contracts. Despite their relevance for
companies, these combination solutions have so far received little attention. We undertake a
systematic literature review to close this research gap and to provide a first comprehensive
overview of this emerging field. We do so by providing a threefold categorization of the different
options for connecting the blockchain and AI.

Keywords: blockchain, artificial intelligence, smart contracts, machine learning, deep learning

Introduction
Among the 21st century’s various disruptive technologies, artificial intelligence (AI) and blockchain
belong to those that stand out in terms of the attention and hype they have received (Salah et al. 2019).
However, bringing together AI and blockchain is a combination of two completely different technologies.
The blockchain is still a very young technology and research field. It was first described in the white paper
Bitcoin: A Peer-to-Peer Electronic Cash System in 2008, although the term "blockchain" was not yet
mentioned (Nakamoto 2008). In comparison, AI has been a well-researched area for several decades,
beginning with the proposal of a first artificial neuron model (McCulloch and Pitts 1943) and Alan
Turing’s highly regarded essay Computing Machinery and Intelligence (1950). Until recently, researchers
only studied blockchain and AI applications in isolation, focusing on their individual application in
different vertical domains and businesses. Nevertheless, connecting AI and blockchain harbors a great
deal of potential. Such a combination is rarely a question of developing completely new applications. Most
of the academic work in this area investigates how one of the two technologies can support the other. For
instance, the blockchain can enable an increase in the transparency of AI systems, which often have the
character of a black box (Castelvecchi 2016). Against this background, the explainable artificial
intelligence (XAI) research field has increasingly become an area of interest for the research community
(Došilović et al. 2018). As an exemplary application, the blockchain can therefore provide new potentials
for increasing transparency (Dillenberger et al. 2019; Sarpatwar et al. 2019). On the other hand, AI can
help overcome some challenges with which that the blockchain, as a new technology, still struggles.
Finally, there is a third category of use cases whose focus is not primarily on one technology supporting
the other. Instead, AI and blockchain are used side by side and unfold their effect through their respective
strengths. Examples include a platform for global employability (Keršič et al. 2020), or an approach that
utilizes blockchain and automated machine learning to provide an automated customer service (Li et al.
2019).

Forty-First International Conference on Information Systems, India 2020

1
 Combining Blockchain and Artificial Intelligence

To date, very little literature deals with blockchain’s role in the AI context (Salah et al. 2019). As our
review of the current literature shows, vice versa also holds true, namely for publications that deal with
AI’s impact on and benefits for blockchain. To this day, the literature lacks comprehensive reviews and
studies on the possibilities that blockchain and AI can develop in cooperation. Initial ideas were already
mentioned in 2014 and 2015 publications. As our literature review shows, since 2017 and 2018 an
increase in research took place. Nevertheless, as Salah et al. (2019) mention, the research field on the
possible combinations of AI and blockchain is still in its infancy. While Salah et al. conducted an extensive
literature research on blockchain’s role in the AI context (Salah et al. 2019), there is as yet, to the best of
our knowledge, no study on blockchain and AI’s combined potential synergies and their subsequent
advantages and benefits. This is surprising, as this combination can be very useful and valuable for a lot of
use-cases. An example is the blockchain’s energy-intensive mining process, which can lead to enormous
energy consumption (Fairley 2017), but AI holds the promise of possible improvements by forecasting the
transaction confirmation time (Singh and Hafid 2020), or through a new consensus procedure (Chen et
al. 2018). In turn, the blockchain can serve as a database to encourage more people to share personal data
(e.g. Zyskind et al. 2015). This data can be highly valuable for companies as a data source for their AI
systems. Given all the benefits for companies, combining AI and blockchain can be considered a very
relevant topic for the IS field. We therefore endeavor to close the research gap by offering a current
overview of the scientific work and approaches in this area. The aim of this work is to provide the latest
state of research on the potential that combined AI and blockchain can have. Our aim also includes
investigating open questions and possible future research directions. Our goal is therefore to find answers
to the following research questions: How can blockchain and artificial intelligence be combined? What are
the possible advantages of such a combination?
The rest of the paper is structured as follows: In section 2, we describe some of both AI and blockchain’s
foundations. Section 3 describes the research method used to conduct the systematic literature review.
Section 4 is divided into three subsections corresponding to the respective possibilities of combining AI
and blockchain. The publications found are presented here, and the current state of research described.
Section 5, the final section, presents a discussion of and concluding remarks about this paper’s results, as
well as providing an outlook on future research.

Foundations
Blockchain
On closer inspection, blockchain is a combination of previously existing technologies. For instance, Diffie
and Hellmann already formulated digital signature and public key cryptography ideas in 1976 (Diffie and
Hellman 1976). Diffie and Hellmann were also the first to recognize the necessity of hash functions for
digital signatures. The first works on hash functions’ analysis and construction were published at the end
of the 1970s by, for instance, Merkle (1979) and Rabin (1978) (Preneel 2010). Further developments later
became part of the blockchain occurred in the 1990s. Haber and Stornetta (1991) recognized the problem
of digital documents being easy to modify. On the basis of hash functions, they proposed two solutions for
digital time-stamping documents so that they cannot be forged (Haber and Stornetta 1991). Satoshi
Nakamoto combined the ideas for group signature schemes (Chaum and van Heyst 1991) and ring
signature schemes (Rivest et al. 2001) with today’s blockchain in 2008 (Nakamoto 2008). Currently , the
blockchain has been widely adopted in various fields and is associated with high expectations (Salah et al.
2019). Some authors argue that the blockchain is the technology with the greatest potential to shape the
next decade’s business world. Tapscott and Tapscott believe that, in this respect, the blockchain is ahead
of other technologies, such as AI, big data, and robotics (Tapscott and Tapscott 2016). According to Iansiti
and Lakhani (2017), as well as Casey and Wong (2017), five basic principles underlie the blockchain
technology:
• Distributed database: The whole database and its complete history are available for each
blockchain party. No single party controls the data and information, and all the parties can verify
the transactions without having to use an intermediary.
• Peer-to-peer transmission: The network peers can communicate with one another without a
central node. This includes the storing and forwarding of information to all other peers.

Forty-First International Conference on Information Systems, India 2020

2
 Combining Blockchain and Artificial Intelligence

• Transparency with pseudonymity: Transactions and their associated values are visible to everyone
within a blockchain network. The single nodes on a blockchain have an alphanumeric address as a
clear identifier. Transactions occur between these blockchain addresses.
• Irreversibility of records: Once a transaction is part of the blockchain, it cannot be altered or
changed, because a transaction is linked to each transaction carried out beforehand. Various
approaches and algorithms ensure transactions’ permanence and correct order.
• Computational logic: The blockchain’s computational logic makes it possible to program on the
blockchain. Users can set up algorithms and rules that automatically trigger transactions between
nodes.

The blockchain’s name is derived from its technical structure, which can also be interpreted as a chain of
blocks with each block connected to the previous blocks by a hash. Until recently, the blockchain was only
used in connection with the Bitcoin, which is the best known project based on the blockchain
(Abbatemarco et al. 2018). Recent research discusses the blockchain’s potential for other application
areas and industries. Examples are the blockchain’s utilization in the context of smart cities (Xie et al.
2019), the internet of things (Christidis and Devetsikiotis 2016), and in the music industry (Baym et al.
2019). Smart contract is another often-mentioned term in connection with the blockchain. This refers to
software that imitates contracts’ behavior or logic, which allows companies to automate contracts’ terms
and conditions. A smart contract can refer to data fields contained in the blockchain (Tapscott und
Tapscott, 2016). Such "contracts" or processes do not require any human interpretation or intervention,
as a computer program executes them (Franco 2015). It should be noted that the blockchain is not a
uniform technology. According to Tasca and Tessone (2019), thousands of blockchain-based projects are
under development worldwide. Some of these projects propose completely new functionalities and
architectures. Consequently, the previously mentioned authors stick to the terms “blockchains” and
“blockchain technologies” (Tasca and Tessone 2019).

Artificial Intelligence
Although mankind has long tried to understand the functioning of intelligence, the term AI was first
coined in 1956 (Russell and Norvig 2016). In recent years, there has been a reemergence of interest in the
field of artificial intelligence among managers and academics (Brock and Wangenheim 2019). Currently,
AI is a broad and thriving field with many practical applications and active research topics (Goodfellow et
al. 2016). Machine learning (ML) technology powers many aspects of modern society: from web searches
to content filtering on social networks, to recommendations on e-commerce websites, and it is
increasingly present in consumer products, such as cameras and smartphones (LeCun et al. 2015). Deep
learning is also among the current trending technologies. Owing to larger datasets and more powerful
computers, deep learning has seen tremendous growth over the last years (Goodfellow et al. 2016). The
architecture of deep learning comprises different modules arranged in multiple layers. Each of these
layers can transform input data and is able to learn. Deep learning has improved the state of the art in
several areas, such as speech recognition, visual object recognition, and object detection (LeCun et al.
2015). Swarm intelligence is another AI discipline concerned with intelligent multi-agent systems’ design
(Blum and Li 2008). This field of research is inspired by swarms from nature like ants or termites, which
have formed a collective behavior. Nowadays, there is a wide range of possible practical applications of
swarm systems. Examples include the transport of large and heavy objects by means of a swarm of mobile
miniature robots (Chen et al. 2013), various applications of swarm robots in the agricultural sector (Emmi
et al. 2014; Yaghoubi et al. 2013), and potential applications for entertainment purposes or toy robots
(Alonso-Mora et al. 2014).

Research Method
According to Kitchenham and Charters, a systematic literature review (SLR) can be used to identify and
evaluate “all available research relevant to a particular research question, or topic area, or phenomenon of
interest” (Kitchenham and Charters 2007). One of this paper’s declared objectives is to provide an as
complete overview of the combination solutions as possible. A SLR was found to be a useful method to
identify all relevant research and scientific work. In conducting the SLR, we followed Kitchenham and
Charters’s (2007) methodology. After precisely defining the research question, we identified the first

Forty-First International Conference on Information Systems, India 2020

3
 Combining Blockchain and Artificial Intelligence

fundamental literature to derive relevant key terms for the search string. We subsequently tested the
string in the used databases to guarantee its functionality. This resulted in the following string used for
the search: “(“blockchain” OR “distributed ledger” OR (“blockchain” AND “smart contract”)) AND
(“intelligent” OR “intelligence” OR “artificial intelligence” OR “machine learning” OR “deep learning” OR
“distributed intelligence” OR “neural network”).” We conducted searches with this string in the following
databases:
• AIS Electronic Library (https://aisel.aisnet.org)
• EBSCO Host (https://search.ebscohost.com)
• Emerald Insight (https://www.emerald.com/insight/)
• Proquest (https://www.proquest.com/)
• Science Direct (https://www.sciencedirect.com/)
• SpringerLink (https://link.springer.com/)
• Web of Science (https://webofscience.com/)
• Wiley Online Library (https://onlinelibrary.wiley.com/)
We followed Mikalef et al. (2018) and Dybå and Dingsøyr’s (2008) approaches to select the identified
studies. After eliminating the duplicates, a total of 1,776 publications remained.

Figure 1. Illustration of the elimination criteria and the resulting

number of publications
Figure 1 illustrates the different process steps to identify the relevant sources. The first step excluded all
publications not published in a peer-reviewed journal or conference. We also excluded white papers, as
they were often rather vague and lacked technical details. The white paper by SingularityNet (2019) was
the only exception to this rule, because identified literature cited it. After this step, 1,435 articles
remained. In the next step, we evaluated each study separately and in more detail. We removed more
publications after reading the titles. This resulted in a total of 1,104 publications. These publications’
abstracts were read to gain a first impression of the exact content. This led to another 769 papers being
eliminated. A more detailed analysis of the remaining 335 articles’ text was necessary in order to classify
the content.
The reading of the 335 articles’ abstracts and entire text led to a particularly large number of articles
identified as unfitting for our study, which resulted in just 28 publications being found relevant. A major
reason for this small number is that trend technologies, such as AI and blockchain, are often mentioned
together. In most cases, however, these technologies are treated separately, while their combined
potentials are only very rarely examined. We did additional backward and forward search searches based
on the remaining publications. During this last step, we found four additional publications not included in

Forty-First International Conference on Information Systems, India 2020

4
 Combining Blockchain and Artificial Intelligence

previous database search. This led to a total of 32 highly relevant articles and publications suitable for
answering our previously defined research questions.

Findings
The identified combination applications can be divided into three main categories. The first category
consists of use-cases and applications that use the blockchain to support or enhance existing AI
applications. Consequently, and in line with previous research, we called this category blockchain for AI.
In contrast, the second category (AI for blockchain) consists of use-cases where AI applications aim to
improve the blockchain. This includes the mining process and smart contracts. Finally, the third category
consists of applications and platforms that use AI and blockchain together without focusing on the one
supporting or leveraging technology through the other one. We decided to call this third category AI with
blockchain. As far as we know, no other research has as yet described this category.

Blockchain for AI
The first main category consists of use cases that aim to support AI methods and techniques by means of
the blockchain. Work dealing with the potential of blockchain for AI has been found most frequently.
These applications can be divided into the following subcategories:
• Blockchain-based data management
• Blockchain-based data marketplaces
• Blockchain-based AI architectures
• Blockchain-enhanced swarm systems
• Increase of transparency through the blockchain
The order of the categories in the list above and in the following is arbitrary and has no bearing on their
priority.

Blockchain-based data management

The huge amount of available data is a driving factor behind the current AI revolution (Dinh and Thai
2018). Using the blockchain to store, manage, and share data has some advantages and potential gains for
AI and ML systems. According to Salah et al., centralized data storage via clouds, data centers, and
clusters is becoming a major bottleneck for highly secure and data-protection-relevant AI applications’
development. Centralized data storage is very vulnerable in terms of data protection and security when it
comes to personal and sensitive data about users, locations, activities, and health records (Salah et al.
2019). Furthermore, there are several ways to manipulate the training data and models of AI systems. The
field of adversarial machine learning is concerned with creating algorithms that are more robust towards
these security challenges. This is a topic of significant relevance, also addressed by the NIST, that
published a first draft for the taxonomy and terminology of adversarial machine learning (Tabassi et al.
2019). Using the blockchain to obtain and store learning data for AI and ML systems holds the promise of
opportunities to counteract the difficulties and dangers mentioned above. This means that data’s secure
storage and provision via the blockchain are a direct advantage for AI and ML systems. Possible gains are
greater security, better quality of learning data, and increased motivation for data creators and owners to
share their data.
Zyskind et al. (2015), who investigated the blockchain’s potential to create a decentralized system for
managing personal data, were responsible for an early approach. Their focus was on mobile applications,
since these collect personal data without the user always being aware of it or being able to control it. These
authors therefore presented a protocol in which a blockchain acts as an automated access control
manager, eliminating the need for third-party control or trust. In contrast to Bitcoin, the transactions in
this system are not of a financial nature, but are used to implement data-related instructions, such as
storage, retrieval, and sharing. In this approach, the blockchain only serves as an instrument for access
control, while the data are not stored on the blockchain (Zyskind et al. 2015). Xia et al. investigate the
potential of a blockchain-based data exchange of healthcare data in cloud environments (Xia et al. 2017).
The dissemination of medical data beyond individual institutions’ protected cloud, poses a serious threat
to patients’ privacy. Xia et al. propose a blockchain-based data sharing framework that adequately

Forty-First International Conference on Information Systems, India 2020

5
 Combining Blockchain and Artificial Intelligence

addresses the access control challenges associated with sensitive data stored in the cloud as a novel
solution. In this approach, like the one described by Zyskind et al., the blockchain does not serve as a
storage medium for the data, but its purpose is instead to enable access or to ensure that medical data can
be provided more securely (Xia et al. 2017).
Shrestha and Vassileva (2018) present a blockchain-based model for collecting research data. This system
should ensure the maintenance of complete and updated research data and a verifiable record of origin.
All accesses to, the common use of, and the sharing of data should also be possible. This should not only
lead to greater transparency for data owners, but also to protecting data against misuse. Users should also
be given additional incentives, for example, digital tokens, to pass their data on to interested data seekers
(Shrestha and Vassileva 2018). Yue et al. (2016) propose an app architecture based on the blockchain
called Healthcare Data Gateway. The aim is to enable patients to easily and securely own, control, and
share their data without compromising privacy. This might lead to new opportunities to improve
healthcare systems’ intelligence while keeping patient data private (Yue et al. 2016). Shafagh et al. (2017)
focus on blockchain-based sharing and storage of data from the Internet of Things (IoT). In this approach,
the blockchain is used as a control layer providing access to the storage. This control layer allows a secure
and resilient access control management (Shafagh et al. 2017). With FairAccess, Ouaddah et al. (2017)
propose a decentralized, pseudonymous, and data protection-oriented authorization management that
uses blockchain technology to manage constrained devices’ access control in the IoT context.

Blockchain-based data marketplaces

Another aspect in this context is the concept of electronic data marketplaces based on a blockchain.
Subramanian (2017) proposes these decentralized marketplaces as a counter-draft to the company-
controlled ones. In a decentralized marketplace, a network of nodes replaces the company responsible for
the marketplace’s proper functioning. This network is responsible for matching buyers and sellers,
enabling transactions and for the infrastructure. The network of nodes therefore provides the same
functionality independently and at the same time as a centralized marketplace would (Subramanian
2017). According to Subramanian, such decentralized marketplaces minimize transaction costs, as they do
not require intermediaries, and buyers can pay sellers directly. Moreover, owing to the fast network
validation, transactions occur immediately, and payment modalities do not delay them. There are also
gains in security, as transaction details can be encrypted, network manipulation is extremely difficult and
cost-intensive, and identities do not need to be revealed on the marketplace (Subramanian 2017).
Özyılmaz, Doğan, and Yurdakul (2018) propose a blockchain-based marketplace for IoT data. Owing to
their ever-increasing number and distribution, IoT devices are increasingly relevant as data generators.
Such a decentralized IoT data platform, based on the blockchain, can offer all its participants various
advantages. According to the mentioned authors, these advantages include not only economic gains, but
also technical and user-related benefits. AI and ML providers are therefore given access to a huge data
pool that was previously not available in this form. This increase in training and test data would lead to AI
and ML systems’ improved performance, which would ultimately also benefit companies or end-
consumers (Özyılmaz et al. 2018).
Mamoshina et al. (2018) present the decentralized marketplace concept for personal health data based on
a blockchain. With the help of such a decentralized marketplace, the authors envisage many opportunities
for discovering drugs, developing biomarkers, and preventive health care. The authors also emphasize the
challenges that data management faces in the health sector. The health sector is specifically challenged to
ensure a high level of data protection and security. Data breaches in health storage systems can be
particularly costly, since various policies threaten reputational damage and high penalties. Mamoshina et
al.’s proposed marketplace is aimed at returning individuals’ control of their personal data, including their
medical records, to them. The marketplace does this by allowing users to upload their data directly into
the system and allowing the use of these data once they have been purchased through the system. In
addition, the marketplace ensures fair tracking of all the data use activities (Mamoshina et al. 2018).
Montes and Goertzel (2019) suggested another approach, the SingularityNET project. This is a platform
for an open AI marketplace on which buyers and sellers can exchange AI services via a blockchain, and AI
agents can trade with one another. The blockchain acts as a basis for a network-internal crypto currency.
The latter tokens are the means of payment for or the medium for the exchange of AI services and for the
internal coordination of AI agents (Montes and Goertzel 2019). Through smart contracts, the AI agents

Forty-First International Conference on Information Systems, India 2020

6
 Combining Blockchain and Artificial Intelligence

will be able to request the performance of AI work, exchange data, and deliver the results (SingularityNet
2019). The various network nodes and participants work together to not only solve AI tasks, as explained
below, but also to build large, decentralized data sets. SingularityNet allows data producers to set usage
restrictions for their data and to receive percentage payments if they are used. Similarly, SingularityNet
participants can track their contributions and their evolution, as well as gain financial benefits when the
data sets and collaborative services grow. By decentralizing ownership and access to records, Montes and
Goertzel see a great potential for SingularityNet regarding simplifying access to and the use of AI
technology, especially for smaller companies. This would counteract the current trend of large technology
giants hoarding very large data sets and smaller companies lacking access to data and the expertise to
prepare these (Montes and Goertzel 2019).

Table 1. Overview of identified benefits of blockchain-based data management and data marketplaces
Higher Increased Control about Increased Easier
Security Privacy own data Transparency regulation
Zyskind et al. 2015 X X X
Xia et al. 2017 X X X
Shrestha and Vassileva 2018 X X X X
Yue et al. 2016 X X X X X
Shafagh et al. 2017 X X X
Ouaddah et al. 2017 X X X
Subramanian 2017 X X
Özyılmaz et al. 2018 X X
Mamoshina et al. 2018 X X X X
Montes and Goertzel 2019 X
Σ 7 8 8 5 2

In the current literature, blockchain-based data management and data marketplaces are the most
investigated blockchain for AI applications. The benefits identified in the literature are manifold, with
increased privacy (e.g., Xia et al. 2017), increased transparency (e.g., Shrestha and Vassileva 2018), and
the possibility of returning data owners’ control of their data to them, mentioned most often(Yue et al.
2016). According to some authors, blockchain’s decentralized architecture will lead to greater security
(Özyılmaz et al. 2018; Shafagh et al. 2017). Table 1 provides an overview of the most often mentioned
advantages and benefits of blockchain-based data management and data marketplaces.

Blockchain-based AI architectures

DeepRing is an architecture that Akhil Goel et al. (2019) propose to protect deep neural networks by
means of the blockchain. Individual blocks of convolutional neural networks (CNNs) are arranged
randomly and contain information about the closest legitimate blocks. An advantage of this architecture is
that DeepRing can register and detect all attacks on CNNs. Tampering at a certain point changes the
respective block’s hash value and that of all subsequent blocks, making changes or attacks visible and
recognizable very quickly and easily. This increases the entire network’s transparency, and that of
individual blocks. The authors used experiments to show that blockchain’s and deep neural networks’
synergy can create tamper-proof models (Goel et al., 2019). Sgantzos and Grigg believe that a blockchain
does not only serve as a basis for data, but as one for an entire artificial intelligence that works with its
own data (Sgantzos and Grigg, 2019). These authors maintain that implementing a swarm of artificial
intelligence agents (AIAs) would form a Church-Turing-Deutsch principle machine. According to the
authors, this would impact different areas, such as the Internet of Things (IoT), financial markets, smart
cities, and personalized medicine (Sgantzos and Grigg, 2019). In a previous publication, Sgantzos already
showed that it is theoretically possible to implement certain cellular automata by means of a blockchain.
Simulating a Naturally Random Generated Mutational System to allow a developed genetic algorithm to

Forty-First International Conference on Information Systems, India 2020

7
 Combining Blockchain and Artificial Intelligence

evolve, should make such an implementation possible. Dedicated coins and tokens will make the
interaction with such a system possible via the blockchain (Sgantzos, 2017).
In the same article, Montes and Goertzel also examine the advantages and gains that SingularityNet
might bring to distributed AI (Montes and Goertzel 2019). In addition to merely exchanging data, AI
agents should also be able to request AI work or provide results by means of smart contracts. The ability
to quickly and dynamically connect or combine different AI systems or agents is also of central
importance. In this way, it should be possible to react adequately to different requests to the network,
which require different capabilities or AI systems. Agents should also be able to develop and train new
and independent AI agents so that the network can be automatically coordinated and further developed.
Similar to the exchange of data, this application will be paid for by means of a blockchain-based token. In
the application that SingularityNet aims at, the blockchain for decentralized AI applications should
provide the combined advantage of facilitating several agents’ coordination with one another. Further,
smart contracts help AI agents interact with one another, with external customers, and when requesting
various tasks or data, as well as payment for them (Montes and Goertzel 2019). However, SingularityNet
is currently not an executable platform and is still under development. Mariani et al. (2017) investigate
different technologies’ influence, including that of the blockchain, on tuple-based coordination systems.
They argue that the blockchain technology’s properties, such as its traceability and transparency, should
further improve and refine tuple-based models’ suitability. These authors see the possibility of managing
interactions between the different involved agents by means of the blockchain and of creating interaction
traces dynamically. It should also be possible to link interaction events by means of transactions and to
provide a kind of accounting of the interactions’ history within the system. This would make it possible to
reliably determine who accessed what based on which previous events or facts, which would be a great
advantage, for example, to later clarify responsibility for a particular event or action. The blockchain can
also be used to monitor the interactions and coordinate the rules. A type of event correlation would be
possible, since the blockchain ensures that distributed transactions are stored and visible in a very
ordered and consistent way; that is, in the correct sequence (Mariani et al. 2017).

Blockchain-enhanced swarm systems

In the context of artificial swarm intelligence, using the blockchain holds the promise of certain
advantages and possible improvements. The blockchain might help increase safety, support distributed
decision making, and differentiate between individual robots’ behavior. It was long assumed that swarm
systems are robust, and that individual robots’ failure has little effect on a swarm’s overall collective
behavior and security (Millard et al. 2013). Bjerknes and Winfield (2013) showed that the latter is not
true, and that the overall system’s reliability decreases with an increasing swarm size. Higgins et al.
(2009) define various potential security gaps and dangers that can arise in the context of swarm systems.
For example, security threats can emerge from insecure communication channels or from manipulated
swarm members. According to Ferrer (2018), the blockchain can provide a reliable peer-to-peer
communication channel for each agent in a swarm, therefore counteracting potential threats,
vulnerabilities, and attacks. In the case of swarm robotics, public-key cryptography allows robots to share
their public keys with other robots for communication purposes. Consequently, each robot in the network
can send information to specific robot addresses by ensuring that only the robot with an appropriate
private key can read the message (Ferrer 2018). The blockchain can also prevent third-party robots from
decrypting information, even if they use the same communication channel. Similarly, a message’s
authorship can be clearly proven when robots use their own private key to encrypt messages (Ferrer
2018). Strobel et al. (2018) present a proposal in their article on how the blockchain can increase the
safety of swarm systems. The authors developed a coordination mechanism that uses a blockchain for
various purposes. Each robot keeps a separate copy of the blockchain and acts as a node and a miner in
the blockchain network. The blockchain serves as a medium for exchanging knowledge, capturing voices,
and applying decision strategies. The coordination functions are realized via a smart contract. This
contract allows the swarm members to perform various functions and activities, including voting or
making decisions. To trigger these functions’ execution, the robots create signed transactions and send
them to the network via the blockchain protocol. In their experiments, the authors showed that it is
possible to identify and exclude malicious or byzantine robots from a swarm (Strobel et al. 2018).
Decision making algorithms are another crucial component in the field of swarm intelligence. They are
used, for instance, for the dynamic allocation and execution of tasks (Das et al. 2011), or to control mobile

Forty-First International Conference on Information Systems, India 2020

8
 Combining Blockchain and Artificial Intelligence

robots’ collective movements (Navarro and Matía 2011). Nguyen et al. (2020) argue that swarm systems’
decentralized nature makes them suitable to be combined with the blockchain. These authors propose a
new, distributed collective decision algorithm for swarm robotics through which robots form a peer-to-
peer network and perform various transactions using blockchain technology. The authors implemented
various collective decision algorithms with and without blockchain participation to perform a
benchmarking exercise. They subsequently carried out tests and compared the results with respect to
various indicators, such as the consensus time and exit probability. Nguyen et al. showed that their
proposed method surpasses other methods without a blockchain (Nguyen et al. 2020).
On the other hand, in the context of swarm robotics, Ferrer (2018) proposes using the blockchain as an
instrument to facilitate decision making and consensus. Once a swarm member is faced with a situation
that requires agreement, it can initiate a special transaction stored in the blockchain and which is
therefore visible to all the other swarm members. Different addresses are linked to this transaction, each
of which represents one of the available decision options. The other swarm members can now vote by
sending a token, corresponding to the option that the specific swarm member selected, to the address.
Agreements can thus be reached quickly, securely, and verifiably by means of, for example, the majority
rule, since all robots can monitor the balance of the addresses involved in the voting process. According to
Ferrer, the described method avoids a training and learning phase for new swarm robots, since all
agreements and all related transactions are part of the blockchain. By downloading the ledger, new robots
automatically obtain the history of all previous decisions and agreements, which enables an automatic
synchronization with the rest of the swarm (Ferrer 2018). However, as the number of transactions
increase, the blockchain also has an increasing amount of data to store. From a certain size onward, it
becomes increasingly difficult to download or save the blockchain (Wagner 2014). This so-called
“bloating” can become an issue if the blockchain is used in the swarm systems context. If large numbers of
robots are used over a long period of time, they can extend the blockchain to such an extent that they can
no longer keep a copy of the entire master book of transactions (Ferrer 2018).
Nishida et al. (2018) address this issue. These authors also see the continuously increasing blockchain as a
challenge for its use in the swarm systems environment. They therefore propose an approach by which the
blockchain’s size and growth can be reduced. This proposal only comprises the storage of hash values in
the blockchain, which are generated from the target data to be exchanged. The size of the transactions
contained in the blockchain is therefore always the same and does not depend on the shared information’s
size. Tests that Nishida et al. carried out showed that the new method reduces the increase in the
blockchain size by 73.0% (Nishida et al. 2018).

Increased transparency

According to Salah et al., AI systems’ decisions that consumers or users find difficult to understand might
lose their value (Salah et al. 2019). A clear audit trail might not only improve the data’s and models’
trustworthiness, but also provide a transparent way of tracking the process of how the AI system came to
its decision (Corea 2019). Dillenberger et al. (2019) assume that the blockchain can increase trust in and
the transparency of AI systems. The authors emphasize that with AI’s increasing importance in everyday
life and in critical business processes, trust in data, models, training processes, and results is of increasing
importance. The blockchain can help track and illustrate a specific AI process at different granularity
levels. The blockchain can also enable a fair evaluation of different stakeholders’ contributions by
capturing the involved parties’ interactions and activities (Dillenberger et al. 2019). This approach is
further elaborated in another publication, in which some of the same authors (Sarpatwar et al. 2019)
present a generic blockchain library to create trust in distributed AI applications and processes. The
library captures the entire distributed AI training process, including the data, intermediate and final
models, processes and dependencies, participants, operations, and relevant metadata (Sarpatwar et al.
2019).

AI for blockchain
In the second main category, the roles of artificial intelligence and blockchain change. This category
consists of scientific works dealing with the blockchain’s support or enhancement through artificial
intelligence methods. This topic has different names: Panetta (2019) calls the combination of the
blockchain and other technologies, such as AI or the internet of things, enhanced blockchain solutions.

Forty-First International Conference on Information Systems, India 2020

9
 Combining Blockchain and Artificial Intelligence

Garimella and Fingar (2018) use the term Blockchain 4.0 when referring to intelligent blockchain
applications. Our research established that this second category has to date received far less scientific
attention than the first one. We divided the identified studies into two categories: the improvement of
smart contracts by means of AI to make them more intelligent, and the improvement of the mining
process.

Intelligent smart contracts

Artificial intelligence’s possible potentials are specifically mentioned in terms of smart contracts. An AI
system could, for example, function as a recommendation system during contract negotiations in supply
chains (Almasoud et al. 2018). Based on archived smart contracts, the AI system could analyze how the
parties negotiated in the past. This would allow the recommendation system to propose an appropriate
language and clauses that would most likely lead to agreement between the different parties involved.
Similarly, the AI can analyze past contracts to identify not yet considered factors in order to integrate
them into future contracts (Almasoud et al. 2018). Additionally, in the smart contracts’ context, an AI
could be programmed to negotiate different conditions that could affect certain goods’ price or quality.
Furthermore, AI systems using supervised or unsupervised learning techniques or methods could
determine the optimal time to trigger or perform a smart contract (Nguyen and Bailey 2018). For
example, a purchase could be made at an optimal time in order to maximize a certain price-performance
ratio. Such a smart contract would also react automatically to price or contract changes, and, based on
this, dynamically change a service or a good’s quality (Nguyen and Bailey 2018).
In addition, AI also has the potential to help in the event of contracts’ failure. In this scenario, an AI
system could help find alternative solutions dynamically and specifically tailored to the parties’ needs
(Nguyen and Bailey 2018). Omohundro (2014) believes that a combination with artificial intelligence
methods would be required to make smart contracts operational in IoT environments. Conventional
smart contracts are often sufficient with regard to processing purely digital transactions. However, as
soon as an interaction with the real and physical world begins, more intelligence and real knowledge of
decision making would probably be required. AI systems are needed to translate information from a
variety of sensors into precise terms to which smart contracts can respond. Similarly, smart contracts that
interact with the real world and lead to physical actions, such as the delivery of objects, should be linked
to human or robotic agents (Omohundro 2014).

Improvement of the mining process

Singh and Hafid (2020) explore the possibility of using machine learning to predict the Ethereum
blockchain’s transaction confirmation time. As the number of transactions increases, the latter authors
see the need for users to know if their transaction will be accepted and how long this is likely to take. They
present three different models that can predict this confirmation time. These three models are each based
on different machine learning methods: Naïve Bayes Classifiers, Random Forests, and Multi-Layer
Perceptrons. Despite offering these approaches, the authors stress that this research field is still in its
infancy and offers much potential for further research (Singh and Hafid 2020).
Chen et al. (2018) also criticize the immature consensus mechanisms currently used in blockchains. They
suggest a completely new consensus mechanism as an alternative, which they call Proof of Artificial
Intelligence (PoAI) (Chen et al. 2018). This new and energy-saving consensus protocol should help ensure
a blockchain network’s decentralization and security. All nodes of a blockchain network are sorted by a
CNN, according to various criteria, such as computing power or security aspects, in a ranking list. Only
the best nodes, which become part of a mining pool, can take over the mining. The relevant mining node
is selected from this mining pool according to a rotation mechanism. This approach has some advantages,
for example, competing for computing power is no longer necessary, which saves electricity and ensures
fairness and decentralization. According to the mentioned authors, the results of the first experiments
with this algorithm are promising (Chen et al. 2018).

Blockchain with AI
Contrary to previous literature, we believe that it is necessary to introduce a third category of use cases in
order to provide a precise classification of the combination potentials. This third category does not focus

Forty-First International Conference on Information Systems, India 2020

10
 Combining Blockchain and Artificial Intelligence

on either AI or blockchain’s support or the enhancement by the other technology. Instead, AI and
blockchain exist side by side and through their combination, create completely new applications. We
therefore call this third category blockchain with AI.
Markopoulos et al. describe an approach using AI and blockchain in combination in the human resources
management context (Markopoulos et al. 2020). This work’s starting point is the Democratic Teaming
Model (DTM) aimed at selecting project personnel democratically. However, this model depends on the
many different skills of the team builder, who is the sole decision maker. Markopoulos et al. elaborate
how both AI and blockchain hold a promise of potential gains for the DTM. By including expert systems,
the organization can obtain recommendations on how teams should be composed. The expert system can
use various types of employee data, such as their interests, experiences, or past activities. The blockchain
can support this further by securing the data feed and transactions to optimize the analytical output.
Besides expert systems, Markopoulos et al. see potentials for other AI technologies, such as machine
learning, pattern recognition, or case-based reasoning (Markopoulos et al. 2020). Keršič et al. (2020)
develop a platform for global employability based on AI and blockchain (Keršič et al. 2020). Their goal is
to create a platform that allows an automatic search and recruitment process. Here, the blockchain is
mainly used to ensure data integrity and to automate the business logic through smart contracts. In Keršič
et al.’s approach, ML, as a part of AI, is used to analyze large amounts of data. ML can help find a suitable
employee for a given task, which would avoid applicants and job offers’ manual screening (Keršič et al.
2020). Arora et al. (2020) propose a combination of deep learning or AI and blockchain to make
collaborative recommender systems safer. The goal of this approach is to make simple control of own data
possible. The mentioned authors believe that the combination of these technologies would be very useful
for those industries in which data privacy is important (Arora et al. 2020).
Ladia (2020) also addresses the issue of companies unable to share data with one another for privacy
reasons. This is a huge disadvantage, because machine learning models benefit from additional training
data. This author presents a blockchain-based implementation allowing the training of machine learning
models without compromising privacy as a solution. In this approach, the blockchain handles the joint
ownership and control of a training machine. This training machine acts as an independent, secure
container that receives training data and untrained models as input. The training machine trains the
respective model internally and returns the trained model as output. To ensure maximum safety, the data
are not visible to any other party and are automatically deleted after the training process (Ladia 2020). Li
et al. (2019) also present an approach based on blockchain and automated machine learning (AutoML) for
an open and automated customer service. The starting point here is data collected by IoT devices during
customer service. These data can be traded in an open, but secure, way with the blockchain. The
blockchain can also be used to ensure that the data are not changed or manipulated. AutoML does the
analysis to reduce dependency on human experts. These authors believe that their approach can
particularly help small and medium-sized enterprises (SMEs). SMEs often do not have the necessary
resources, since machine learning methods’ development is often time-consuming as well as labor and
knowledge intensive. AutoML in combination with the blockchain should allow this process to be
automated to make it more accessible (Li et al. 2019).
Mylrea (2019) investigates how blockchain, AI, and machine learning together could form distributed
autonomous energy organizations (DAEOs). Such a DAEO would serve as a basis of a distributed
autonomous system for trading energy. This would have much potential, such as improving both the
security and the speed, as well as reducing the need for third parties. Autonomous smart contracts could
use autonomous agents to automatically exchange values or services. In this context, the blockchain can
also function as a secure storage medium, through which the data is cryptographically signed and securely
stored in a distributed ledger. This would allow the blockchain to be used, especially when the DAEO’s AI
requires sensitive data. Mylrea also takes a critical look at the blockchain, showing when it could make a
meaningful contribution, but also when a conventional database would be more suitable. However,
Mylrea concludes that a DAEO, equipped with AI and blockchain, would improve the electrical
infrastructure’s safety, efficiency, and resilience (Mylrea 2019).

Discussion and Conclusion

Both artificial intelligence and blockchain are technologies which harbor great potential for future usage.
The same probably holds true regarding the solutions and applications in which blockchain and AI are

Forty-First International Conference on Information Systems, India 2020

11
 Combining Blockchain and Artificial Intelligence

used in combination. Such combination solutions can be divided into three different groups: on the one
hand, there are solutions in which the blockchain is used to support AI systems, or to address the
associated difficulties (blockchain for AI). In contrast, applications in the AI for blockchain field are
intended to help overcome blockchain-related issues with the help of AI and machine learning solutions.
Besides these two categories, which the literature has already described, our paper introduces a third
category, blockchain with AI. In the applications of this new category, blockchain and AI coexist. One
technology’s support of the other does not apply to applications in this third category. We considered it
necessary to introduce this third category in order to have a comprehensive overview of all the blockchain
and AI combination solutions. By providing an overview of the current scientific research and
publications, our literature review contributes to the research on how blockchain and AI can be combined.
The categorization of the papers that we identified in this regard is intended to be a first attempt to order
and fully categorize the existing research. The existing studies only concentrate on individual areas, for
example, the blockchain for AI (Salah et al. 2019). Our paper is, on the other hand, aimed at providing a
complete overview of all the categories of possible combinations, which also include a description of the
possible application potentials or gains, that can arise from an AI and blockchain combination. We believe
that we succeeded in providing interested researchers with an initial overview of this emerging research
field.
In view of future technological developments’ unpredictability, this paper’s results only represent the
current state of the art in research. One could, however assume that the tripartitioning of the research
field (blockchain for AI, AI for blockchain, blockchain and AI), which this paper describes for the first
time, will continue to be a meaningful form of this field’s structuring in the future. The categories of
combination solutions compiled here, as well as their disadvantages and weaknesses, are also likely to
retain their relevance and become the subject of future projects and research. However, the completeness
of the applications and application categories described here cannot be guaranteed in the long term, since
this study only investigates scientific publications and articles. Owing to technology’s fast development,
there might be use-cases in practice that research has not yet considered. Consequently, a closer
investigation of the grey literature, white papers, and use-cases the practice might provide further
insights. In addition, AI and blockchain’s combination applications cannot be viewed in isolation. Instead,
both AI and blockchain are continuously subject to new research results and developments. These new
results and progress might have a direct influence on the combined applications field. It might, for
instance become possible for each of the two technologies to address the other’s disadvantages and
weaknesses even further in the future. This would lead to the development of further applications in the
areas of blockchain for AI and AI for blockchain. Furthermore, it is conceivable that some of AI and
blockchain’s currently existing disadvantages might be solved without the intervention of one of the two.
This might make some combination solutions obsolete. It should be remembered that the combination of
AI/ML and blockchain does not represent a combination of two clearly defined technologies. Instead, they
should be understood as generic terms that cover many different techniques and methods. The latter
creates a great deal of complexity and an enormous amount of potentially conceivable configurations and
combinations, besides giving rise to a multitude of further research questions. These questions could be
used to investigate the suitability of different AI and ML methods, as well as the different blockchain
variants, for combination solutions. Salah et al. (2019) provide an initial list of future and open research
questions. Completely new technologies, such as quantum computing, which could have a direct influence
on the blockchain are also of interest (Kiktenko et al., 2018; Rodenburg and Pappas, 2017). We conclude
that many interactions with other technologies and disciplines are possible. It is therefore likely that the
research on combining AI and blockchain will undergo dynamic developments in the future.

Forty-First International Conference on Information Systems, India 2020

12
 Combining Blockchain and Artificial Intelligence

Appendix – Full list of investigated publications

Table 2. Full list of investigated publications

ID Author(s) Year Category Application
01 Shrestha and Vassileva 2018 Blockchain for AI Blockchain-based data management
02 Shafagh et al. 2017 Blockchain for AI Blockchain-based data management
03 Ouaddah et al. 2017 Blockchain for AI Blockchain-based data management
04 Xia et al. 2017 Blockchain for AI Blockchain-based data management
05 Yue et al. 2016 Blockchain for AI Blockchain-based data management
06 Zyskind et al. 2015 Blockchain for AI Blockchain-based data management
07 Montes and Goertzel 2019 Blockchain for AI Blockchain-based data marketplace
08 Mamoshina et al. 2018 Blockchain for AI Blockchain-based data marketplace
09 Özyılmaz et al. 2018 Blockchain for AI Blockchain-based data and AI marketplace
10 Mariani et al. 2017 Blockchain for AI Blockchain-based decentralized intelligence
11 Nguyen et al. 2020 Blockchain for AI Blockchain-based swarm system support
12 Ferrer 2018 Blockchain for AI Blockchain-based swarm system support
13 Nishida et al. 2018 Blockchain for AI Blockchain-based swarm system support
14 Goel et al. 2019 Blockchain for AI Blockchain-based AI infrastructure
15 Sgantzos and Grigg 2019 Blockchain for AI Blockchain-based AI infrastructure
16 Sgantzos 2017 Blockchain for AI Blockchain-based AI infrastructure
17 Dillenberger et al. 2019 Blockchain for AI Blockchain-based increase of transparency
18 Sarpatwar et al. 2019 Blockchain for AI Blockchain-based increase of transparency
19 Almasoud et al. 2018 AI for blockchain AI-supported smart contracts
20 Nguyen and Bailey 2018 AI for blockchain AI-supported smart contracts
21 Omohundro 2014 AI for blockchain AI-supported smart contracts
22 Singh and Hafid 2020 AI for blockchain AI-enhanced mining process
23 Chen et al. 2018 AI for blockchain AI-enhanced mining process
24 Markopoulos et al. 2020 Blockchain with AI Parallel use of AI and blockchain
25 Keršič et al. 2020 Blockchain with AI Parallel use of AI and blockchain
26 Arora et al. 2020 Blockchain with AI Parallel use of AI and blockchain
27 Ladia 2020 Blockchain with AI Parallel use of AI and blockchain
28 Li et al. 2019 Blockchain with AI Parallel use of AI and blockchain
29 Mylrea 2019 Blockchain with AI DAEO
30 Salah et al. 2019 Blockchain for AI General discussion
31 Corea 2019 General discussion General discussion
32 Dinh and Thai 2018 General discussion General discussion

Forty-First International Conference on Information Systems, India 2020

13
 Combining Blockchain and Artificial Intelligence

References
Abbatemarco, N., Rossi, L. M. de, and Salviotti, G. 2018. “An econometric model to estimate the value of a
cryptocurrency network. The Bitcoin case,” Research Papers 164.
Almasoud, A. S., Eljazzar, M. M., and Hussain, F. 2018. “Toward a Self-Learned Smart Contracts,” in 15th
International Conference on e-Business Engineering: ICEBE 2018: Proceedings, Xi'an, China. 12.
Oktober - 14. Oktober, Los Alamitos, California: IEEE Computer Society, Conference Publishing
Services, pp. 269-273.
Alonso-Mora, J., Siegwart, R., and Beardsley, P. 2014. “Human - robot swarm interaction for
entertainment,” in HRI'14: Proceedings of the 2014 ACM/IEEE International Conference on Human-
Robot Interaction, G. Sagerer, M. Imai, T. Belpaeme and A. Thomaz (eds.), Bielefeld, Germany. 3.
März - 6. März, ACM Press, p. 98.
Arora, M., Chopra, A. B., and Dixit, V. S. 2020. “An Approach to Secure Collaborative Recommender
System Using Artificial Intelligence, Deep Learning, and Blockchain,” in Intelligent communication,
control and devices: Proceedings of ICICCD 2018, S. Choudhury, R. Mishra, R. G. Mishra and A.
Kumar (eds.), Singapore: Springer, pp. 483-495.
Baym, N., Swartz, L., and Alarcon, A. 2019. “Convening Technologies: Blockchain and the Music
Industry,” International Journal of Communication (13), pp. 402-421.
Bjerknes, J. D., and Winfield, A. F. T. 2013. “On Fault Tolerance and Scalability of Swarm Robotic
Systems,” in Distributed Autonomous Robotic Systems: The 10th International Symposium, A.
Martinoli, F. Mondada, N. Correll, G. Mermoud, M. Egerstedt, M. A. Hsieh, L. E. Parker and K. Støy
(eds.), Berlin, Heidelberg: Springer Berlin Heidelberg; Imprint; Springer, pp. 431-444.
Blum, C., and Li, X. 2008. “Swarm Intelligence in Optimization,” in Swarm intelligence: Introduction
and Applications, C. Blum and D. Merkle (eds.), Berlin, Germany: Springer, pp. 43-86.
Brock, J. K.-U., and Wangenheim, F. von 2019. “Demystifying AI: What Digital Transformation Leaders
Can Teach You about Realistic Artificial Intelligence,” California Management Review (61:4), pp.
110-134 (doi: 10.1177/1536504219865226).
Casey, M. J., and Wong, P. 2017. Global Supply Chains Are About to Get Better, Thanks to Blockchain.
https://hbr.org/2017/03/global-supply-chains-are-about-to-get-better-thanks-to-blockchain.
Accessed 30 September 2019.
Castelvecchi, D. 2016. “Can we open the black box of AI?” Nature (538:7623), pp. 20-23.
Chaum, D., and van Heyst, E. 1991. “Group Signatures,” in Advances in cryptology - EUROCRYPT '91:
Workshop on the Theory and Application of Cryptographic Techniques, Brighton, UK, April 8 - 11,
1991 ; proceedings, D. W. Davies (ed.), Berlin: Springer, pp. 257-265.
Chen, J., Duan, K., Zhang, R., Zeng, L., and Wang, W. 2018. “An AI Based Super Nodes Selection
Algorithm in BlockChain Networks,”
Chen, J., Gauci, M., and Gross, R. 2013. “A strategy for transporting tall objects with a swarm of miniature
mobile robots,” in 2013 IEEE International Conference on Robotics and Automation (ICRA):
Proceedings, Karlsruhe, Germany. 6. Mai - 10. Mai, IEEE, pp. 863-869.
Christidis, K., and Devetsikiotis, M. 2016. “Blockchains and Smart Contracts for the Internet of Things,”
IEEE Access (4), pp. 2292-2303.
Corea, F. 2019. Applied Artificial Intelligence: Where AI Can Be Used In Business, Cham: Springer.
Das, G. P., McGinnity, T. M., Coleman, S. A., and Behera, L. 2011. “A fast distributed auction and
consensus process using parallel task allocation and execution,” in 2011 IEEE/RSJ International
Conference on Intelligent Robots and Systems, I. Staff (ed.), San Francisco, CA, USA. 25. September -
30. September, IEEE, pp. 4716-4721.
Diffie, W., and Hellman, M. 1976. “New directions in cryptography,” IEEE Transactions on Information
Theory (22:6), pp. 644-654.
Dillenberger, D. N., Novotny, P., Zhang, Q., Jayachandran, P., Gupta, H., Hans, S., Verma, D.,
Chakraborty, S., Thomas, J. J., Walli, M. M., Vaculin, R., and Sarpatwar, K. 2019. “Blockchain
analytics and artificial intelligence,” IBM Journal of Research and Development (63:2/3), 1-14.
Dinh, T. N., and Thai, M. T. 2018. “AI and Blockchain: A Disruptive Integration,” Computer (51:9), pp.
48-53 (doi: 10.1109/MC.2018.3620971).
Došilović, F. K., Brčić, M., and Hlupić, N. 2018. “Explainable artificial intelligence: A survey,” in 41st
International Convention on Information and Communication Technology, Electronics and
Microelectronics (MIPRO): Proceedings, Opatija, Kroatien. 21. Mai - 25. Mai, IEEE, pp. 210-215.

Forty-First International Conference on Information Systems, India 2020

14
 Combining Blockchain and Artificial Intelligence

Dybå, T., and Dingsøyr, T. 2008. “Empirical studies of agile software development: A systematic review,”
Information and Software Technology (50:9-10), pp. 833-859 (doi: 10.1016/j.infsof.2008.01.006).
Emmi, L., Gonzalez-de-Soto, M., Pajares, G., and Gonzalez-de-Santos, P. 2014. “New trends in robotics
for agriculture: integration and assessment of a real fleet of robots,” The Scientific World Journal
(2014).
Fairley, P. 2017. “Blockchain world - Feeding the blockchain beast if bitcoin ever does go mainstream, the
electricity needed to sustain it will be enormous,” IEEE Spectrum (54:10), pp. 36-59 (doi:
10.1109/MSPEC.2017.8048837).
Ferrer, E. C. 2018. “The Blockchain: A New Framework for Robotic Swarm Systems,” in Proceedings of
the Future Technologies Conference (FTC) 2018: Volume 2, K. Arai, R. Bhatia and S. Kapoor (eds.),
Vancouver, BC, Canada. 15. November - 16. November, pp. 1037-1058.
Franco, P. 2015. Understanding Bitcoin: Cryptography, Engineering and Economics, Chichester, West
Sussex: Wiley.
Garimella, K., and Fingar, P. 2018. AI+Blockchain: A brief guide for gamechangers, Tampa: Meghan-
Kiffer Press.
Goel, Akhil, Agarwal, A., Vatsa, M., Singh, R., and Ratha, N. 2019. “DeepRing: Protecting Deep Neural
Network with Blockchain,” in The IEEE Conference on Computer Vision and Pattern Recognition
(CVPR) Workshops, Long Beach, California. June 16-20.
Goodfellow, I., Bengio, Y., and Courville, A. 2016. Deep learning, Cambridge, Massachusetts, London,
England: MIT Press.
Haber, S., and Stornetta, W. S. 1991. “How to time-stamp a digital document,” Journal of Cryptology
(3:2), pp. 99-111.
Higgins, F., Tomlinson, A., and Martin, K. M. 2009. “Survey on Security Challenges for Swarm Robotics,”
in Proceedings, the Fifth International Conference on Autonomic and Autonomous Systems, R. C.
Calinescu (ed.), Valencia, Spain. 20. April - 25. April, Los Alamitos, Calif.: IEEE Computer Society, pp.
307-312.
Iansiti, M., and Lakhani, K. R. 2017. “The truth about blockchain,” Harvard Business Review (95:1), pp.
118-127.
Keršič, V., Štukelj, P., Kamišalić, A., Karakatić, S., and Turkanović, M. 2020. “A Blockchain- and AI-based
Platform for Global Employability,” in Blockchain and applications: International congress, J. Prieto,
A. K. Das, S. Ferretti, A. Pinto and J. M. Corchado (eds.), Cham, Switzerland: Springer, pp. 161-168.
Kitchenham, B., and Charters, S. 2007. “Guidelines for performing Systematic Literature Reviews in
Software Engineering: Version 2.3,”
Ladia, A. 2020. “Privacy Centric Collaborative Machine Learning Model Training via Blockchain,” in
Blockchain and applications: International congress, J. Prieto, A. K. Das, S. Ferretti, A. Pinto and J.
M. Corchado (eds.), Cham, Switzerland: Springer, pp. 62-70.
LeCun, Y., Bengio, Y., and Hinton, G. 2015. “Deep learning,” Nature (521:7553), pp. 436-444.
Li, Z., Guo, H., Wang, W. M., Guan, Y., Barenji, A. V., Huang, G. Q., McFall, K. S., and Chen, X. 2019. “A
Blockchain and AutoML Approach for Open and Automated Customer Service,” IEEE Transactions
on Industrial Informatics (15:6), pp. 3642-3651 (doi: 10.1109/TII.2019.2900987).
Mamoshina, P., Ojomoko, L., Yanovich, Y., Ostrovski, A., Botezatu, A., Prikhodko, P., Izumchenko, E.,
Aliper, A., Romantsov, K., Zhebrak, A., Ogu, I. O., and Zhavoronkov, A. 2018. “Converging blockchain
and next-generation artificial intelligence technologies to decentralize and accelerate biomedical
research and healthcare,” Oncotarget (9:5), pp. 5665-5690.
Mariani, S., Omicini, A., and Ciatto, G. 2017. “Novel Opportunities for Tuple-based Coordination: XPath,
the Blockchain, and Stream Processing,” in 18th Workshop "From Objects to Agents": Proceedings of
the 18th Workshop "From Objects to Agents", P. de Meo, M. N. Postorino and D. Rosaci (eds.),
Scilla, Italien. 15. Juni - 16. Juni, pp. 61-64.
Markopoulos, E., Kirane, I. S., Balaj, D., and Vanharanta, H. 2020. “Artificial Intelligence and Blockchain
Technology Adaptation for Human Resources Democratic Ergonomization on Team Management,” in
Human Systems Engineering and Design II, T. Ahram, W. Karwowski, S. Pickl and R. Taiar (eds.),
Cham: Springer International Publishing, pp. 445-455.
McCulloch, W. S., and Pitts, W. 1943. “A logical calculus of the ideas immanent in nervous activity,” The
Bulletin of Mathematical Biophysics (5:4), pp. 115-133 (doi: 10.1007/BF02478259).
Merkle, R. C. 1979. Secrecy, authentication, and public key systems: Doctoral Dissertation,
Standford, CA, USA.

Forty-First International Conference on Information Systems, India 2020

15
 Combining Blockchain and Artificial Intelligence

Mikalef, P., Pappas, I. O., Krogstie, J., and Giannakos, M. 2018. “Big Data Analytics Capabilities: A
Systematic Literature Review and Research Agenda,” Information Systems and e-Business
Management (16:3), pp. 547-578 (doi: 10.1007/s10257-017-0362-y).
Millard, A. G., Timmis, J., and Winfield, A. F. T. 2013. “Towards Exogenous Fault Detection in Swarm
Robotic Systems,” in Towards autonomous robotic systems: 14th Annual Conference, TAROS 2013
Oxford, UK, August 28–30, 2013 Revised Selected Papers, A. Natraj, S. Cameron, C. Melhuish and M.
Witkowski (eds.), Oxford, Großbritannien. 28. August - 30. August, Cham: Springer, pp. 429-430.
Montes, G. A., and Goertzel, B. 2019. “Distributed, decentralized, and democratized artificial intelligence,”
Technological forecasting and social change (141), pp. 354-358.
Mylrea, M. 2019. “Chapter 12 - Distributed Autonomous Energy Organizations: Next-Generation
Blockchain Applications for Energy Infrastructure,” in Artificial intelligence for the Internet of
everything, W. Lawless, R. Mittu, D. Sofge, I. S. Moskowitz and S. Russell (eds.), London, United
Kingdom, San Diego, CA: Academic Press, an imprint of Elsevier, pp. 217-239.
Nakamoto, S. 2008. Bitcoin: A Peer-to-Peer Electronic Cash System. https://bitcoin.org/bitcoin.pdf.
Accessed 24 October 2019.
Navarro, I., and Matía, F. 2011. “A framework for the collective movement of mobile robots based on
distributed decisions,” Robotics and Autonomous Systems (59:10), pp. 685-697.
Nguyen, H., and Bailey, S. 2018. “Use of Artificial Intelligence for Smart Contracts and Blockchains,”
FinTechLaw Report: E-Banking, Payments and Commerce in the Mobile World (20:2), pp. 1-7.
Nguyen, T. T., Hatua, A., and Sung, A. H. 2020. “Blockchain Approach to Solve Collective Decision
Making Problems for Swarm Robotics,” in Blockchain and applications: International congress, J.
Prieto, A. K. Das, S. Ferretti, A. Pinto and J. M. Corchado (eds.), Cham, Switzerland: Springer, pp.
118-125.
Nishida, Y., Kaneko, K., Sharma, S., and Sakurai, K. 2018. “Suppressing Chain Size of Blockchain-Based
Information Sharing for Swarm Robotic Systems,” in 2018 Sixth International Symposium on
Computing and Networking Workshops: Proceedings, Takayama, Japan. 27. November -
30. November, Los Alamitos, CA: IEEE Computer Society, Conference Publishing Services, pp. 524-
528.
Omohundro, S. 2014. “Cryptocurrencies, smart contracts, and artificial intelligence,” AI Matters (1:2), pp.
19-21.
Ouaddah, A., Elkalam, A. A., and Ouahman, A. A. 2017. “Towards a Novel Privacy-Preserving Access
Control Model Based on Blockchain Technology in IoT,” in Europe and MENA Cooperation Advances
in Information and Communication Technologies, Á. Rocha, M. Serrhini and C. Felgueiras (eds.),
Cham: Springer International Publishing; Imprint; Springer, pp. 523-533.
Özyılmaz, K. R., Doğan, M., and Yurdakul, A. 2018. “IDMoB: IoT Data Marketplace on Blockchain,” in
2018 Crypto Valley Conference on Blockchain Technology: Proceedings, Zug. 20. Juni - 22. Juni, Los
Alamitos, California, Washington, Tokyo: Conference Publishing Services, IEEE Computer Society,
pp. 11-19.
Panetta, K. 2019. The CIO’s Guide to Blockchain. https://www.gartner.com/smarterwithgartner/the-cios-
guide-to-blockchain/. Accessed 26 August 2020.
Preneel, B. 2010. “The First 30 Years of Cryptographic Hash Functions and the NIST SHA-3
Competition,” in Topics in cryptology: The Cryptographers' Track at the RSA Conference 2010, San
Francisco, CA, USA, March 1-5, 2010. Proceedings, J. Pieprzyk (ed.), Berlin: Springer, pp. 1-14.
Rabin, M. O. 1978. “Digitalized Signatures,” in Foundations of secure computation, R. A. DeMillo (ed.),
New York: Academic Press, pp. 155-166.
Russell, S. J., and Norvig, P. 2016. Artificial intelligence: A modern approach, Boston, Columbus,
Indianapolis: Pearson.
Salah, K., Rehman, M. H. U., Nizamuddin, N., and Al-Fuqaha, A. 2019. “Blockchain for AI: Review and
Open Research Challenges,” IEEE Access (7), pp. 10127-10149.
Sarpatwar, K., Vaculin, R., Min, H., Su, G., Heath, T., Ganapavarapu, G., and Dillenberger, D. 2019.
“Towards Enabling Trusted Artificial Intelligence via Blockchain,” in Policy-Based Autonomic Data
Governance, S. Calo, E. Bertino and D. Verma (eds.), SPRINGER NATURE, pp. 137-153.
Shafagh, H., Burkhalter, L., Hithnawi, A., and Duquennoy, S. 2017. “Towards Blockchain-based Auditable
Storage and Sharing of IoT Data,” in Proceedings of the 2017 on Cloud Computing Security
Workshop, B. Thuraisingham (ed.), Dallas, Texas, USA. 3. November, New York, NY: ACM, pp. 45-50.

Forty-First International Conference on Information Systems, India 2020

16
 Combining Blockchain and Artificial Intelligence

Shrestha, A. K., and Vassileva, J. 2018. “Blockchain-Based Research Data Sharing Framework for
Incentivizing the Data Owners,” in BLOCKCHAIN - ICBC 2018: First international conference, S.
Chen, H. Wang and L.-J. Zhang (eds.), Seattle, WA, USA. 25. Juni - 30. Juni, Springer, pp. 259-266.
Singh, H. J., and Hafid, A. S. 2020. “Prediction of Transaction Confirmation Time in Ethereum
Blockchain Using Machine Learning,” in Blockchain and applications: International congress, J.
Prieto, A. K. Das, S. Ferretti, A. Pinto and J. M. Corchado (eds.), Cham, Switzerland: Springer, pp.
126-133.
SingularityNet 2019. White Paper 2.0: SingularityNET - A Decentralized, Open Market and Network for
AIs. https://public.singularitynet.io/whitepaper.pdf. Accessed 18 October 2019.
Strobel, V., Ferrer, E. C., and Dorigo, M. 2018. “Managing Byzantine Robots via Blockchain Technology in
a Swarm Robotics Collective Decision Making Scenario scenario,” in Proceedings of the 17th
International Conference on Autonomous Agents and MultiAgent Systems, E. Andre, S. Koenig, M.
Dastani and G. Sukthankar (eds.), Stockholm, Schweden. 10. Juli - 15. Juli, 3237383: International
Foundation for Autonomous Agents and Multiagent Systems, pp. 541-549.
Subramanian, H. 2017. “Decentralized blockchain-based electronic marketplaces,” Communications of
the ACM (61:1), pp. 78-84.
Tabassi, E., Burns, K. J., Hadjimichael, M., Molina-Markham, A. D., and Sexton, J. T. 2019. A Taxonomy
and Terminology of Adversarial Machine Learning: Draft NISTIR 8269.
https://nvlpubs.nist.gov/nistpubs/ir/2019/NIST.IR.8269-draft.pdf. Accessed 2 September 2020.
Tapscott, D., and Tapscott, A. 2016. The Impact of the Blockchain Goes Beyond Financial Services.
https://hbr.org/2016/05/the-impact-of-the-blockchain-goes-beyond-financial-services. Accessed 26
October 2019.
Tasca, P., and Tessone, C. J. 2019. “A Taxonomy of Blockchain Technologies: Principles of Identification
and Classification,” Ledger (4), pp. 1-39.
Turing, A. M. 1950. “Computing Machinery and Intelligenc,” Mind (LIX:236), pp. 433-460.
Wagner, A. 2014. Ensuring Network Scalibility: How to Fight Blockchain Bloat.
https://bitcoinmagazine.com/articles/how-to-ensure-network-scalibility-fighting-blockchain-bloat-
1415304056. Accessed 23 September 2019.
Xia, Q., Sifah, E., Smahi, A., Amofa, S., and Zhang, X. 2017. “BBDS: Blockchain-Based Data Sharing for
Electronic Medical Records in Cloud Environments,” Information (8:2), pp. 44-59.
Xie, J., Tang, H., Huang, T., Yu, F. R., Xie, R., Liu, J., and Liu, Y. 2019. “A Survey of Blockchain
Technology Applied to Smart Cities: Research Issues and Challenges,” IEEE Communications
Surveys & Tutorials (21:3), pp. 2794-2830 (doi: 10.1109/COMST.2019.2899617).
Yaghoubi, S., Akbarzadeh, N. A., Bazargani, S. S., Bazargani, S. S., and Bamizan, M. 2013. “Autonomous
Robots for Agricultural Tasks and Farm ssignment and Future Trends in Agro Robots,” International
Journal of Mechanical & Mechatronics Engineering (13:3).
Yue, X., Wang, H., Jin, D., Li, M., and Jiang, W. 2016. “Healthcare Data Gateways: Found Healthcare
Intelligence on Blockchain with Novel Privacy Risk Control,” Journal of medical systems (40:10), p.
218.
Zyskind, G., Nathan, O., and Pentland, A. 'S.' 2015. “Decentralizing Privacy: Using Blockchain to Protect
Personal Data,” in 2015 IEEE Security and Privacy Workshops (SPW), San Jose, CA, USA. 21. Mai -
22. Mai, Piscataway, NJ: IEEE, pp. 180-184.

Forty-First International Conference on Information Systems, India 2020

17

View publication stats