Coping with Artificial Intelligence Ethical Dilemma and Ethical

Position Choices?

Sylvie Gerbaix

, Sylvie Michel

and Marc Bidan

Laboratoire MRM, Montpellier University, Aix-Marseille University, France

Univ. Bordeaux, IRGO, UR 4190, F-33000 Bordeaux, France

LEMNA, Nantes Université, France

Keywords: Artificial Intelligence (AI), Algorithmic System, Ethical Theory, Bias, Responsibility, Transparency.

Abstract: The aim of this conceptual article is to demonstrate that proposing measures, actions, and decisions to

improve the ethics of Artificial Intelligence (AI) depends on the ethical theoretical position chosen. To

achieve this, we proceeded in two stages. Firstly, we characterized and synthesized three different ethical

issues posed by AI. Secondly, we selected two main ethical positions proposed by philosophical literature.

Finally, we showed that the choice of an ethical theoretical position for each category of ethical issues of AI

leads to different decisions. We demonstrated that for each category of ethical problems, the ethical

decisions and their consequences differ depending on the ethical theory chosen. The value of this paper is to

highlight that the literature on AI ethics often neglects the implications of choosing an ethical position. In

order to attempt to solve ethical issues, it is necessary to reach agreements and have discussions that take

into account the different ethical theoretical positions and their consequences in terms of decision-making.

https://orcid.org/0009-0005-3544-2399

https://orcid.org/0000-0002-8175-9996

https://orcid.org/0000-0003-1739-5697

1 INTRODUCTION

Today, AI helps us in selecting footage, music,

friends, and partners (Milano et al., 2020). It also

supports institutions in making legal decisions,

maintaining public order, and helps doctors in

providing a diagnosis (Obermeyer, 2019), traders in

trading (Aggarwal, 2021), and armies in using killer

robots to achieve their goals. Numerous areas seem

to be under the yoke of AI capacities, including

logistics, health, education, research, defense,

banking, agri-food, culture, leisure, social, and

professional networks. Since algorithms are at the

heart of human relations and exchanges, questions

relating to uses (and misuses), benefits (and limits)

have become crucial. Faced with this surge of

artificial intelligence, the ethical question is urgent.

Artificial intelligence covers a wide range of

research and computer applications, including

machine learning, computer vision, knowledge

representation, language processing, and decision

support. With the notion of AI, we propose that it is

not the algorithm alone that can be problematic, but

rather its embeddedness in a system, a set of actors,

power norms, and complexity (Neyland, 2016;

Seaver, 2017). Thus, in this contribution, we

consider AI not as purely technical objects, but as

technical systems embedded in culture(s) and which

can be seen, used, and approached from different

perspectives (legal, technological, cultural, social). It

is a technical construction that is both deeply social

and cultural. It does not escape social and cultural

construction like all other tools, neither in its

development, nor in the inputs (data), nor in the

interpretation of the results (output), nor in its use.

Some authors (Hamet & Michel, 2018) have

shown that the ethical questions that arise in

information systems are specific to this field, in the

sense that the ethical dilemmas posed by AIs do not

arise, or do not arise in the same way in other fields.

However, even if some questions appear to be

specific to AI, it seems necessary to have a

theoretical framework to study them and provide

answers.

382

Gerbaix, S., Michel, S. and Bidan, M.

Coping with Artiﬁcial Intelligence Ethical Dilemma and Ethical Position Choices?.

DOI: 10.5220/0012726000003690

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 26th International Conference on Enterprise Information Systems (ICEIS 2024) - Volume 1, pages 382-388

ISBN: 978-989-758-692-7; ISSN: 2184-4992

Numerous articles today attempt to analyze the

main ethical problems linked to AI by proposing, for

the most part, courses of action, standards, and codes

to be put in place (Mittelstadt et al., 2016; Berreby et

al., 2017; Anderson et Anderson, 2018; Yu et al.,

2018; Buhmann et al., 2020; Tsamados et al., 2021).

However, we note that these studies are not part of a

theoretical ethical current. We would like to show

that when faced with an ethical problem raised by

AI, the recommendations and decisions may not be

the same depending on the initial theoretical ethical

position. To this end, we will first present two main

theoretical ethical currents that we consider

important: Kant's deontological ethics and Hegel's

consequentialist ethics. A synthesis of the ethical

questions raised by AI is then presented. In a third

point, we confront them with the two theoretical

ethical positions. We then show that the decisions to

be taken may differ. Thus, we show that the norms

to be put in place in the face of the ethical problems

of AI will only be effective if a theoretical ethical

position is chosen ex ante.

2 SYNTHESIS OF ARTIFICIAL

INTELLIGENCE (AI) ETHICAL

ISSUES IN THREE THEMES

2.1 AI and Responsibility

Here we need to point to De George (2009) who

highlights an unfortunate trend of giving up the

assignment of responsibilities when it comes to

Information Systems or algorithms, a renunciation

that comes from two sources. The first is what De

George (2009) refers to as the myth of amoral

Information and communication technology (ICT).

This myth amounts to limiting ICT to their technical

aspect and to considering that machines cannot,

obviously, be held responsible for the consequences

of their use. There is an avoidance of responsibility

through “the computer said so” type of denial

(Karppi, 2018). The second source of dilution of

responsibility comes from a split, in decision-

making, between the developers, who believe they

are fulfilling their duty by strictly respecting the

requests they receive, and the management, who

does not consider itself responsible for the

technology-related flaws. However, these developers

are never face to face with the stakeholders which

creates a new ethical problem linked to the distance

in decision-making (Rubel et al., 2019). This issue

of responsibility is very current, particularly with the

development of AI (Martin, 2019; Yu et al., 2019,

Wieringa, 2020).

Faced with this dilution of responsibilities, some

authors (Chatterjee et al., 2009; Light and McGrath,

2010) call for the adoption of a disclosing ethic

approach considering the algorithms as an actor in

its own right, and aiming to reveal the ethical

questions posed by their design, and not just their

use. Machines should be able to make ethical

decisions using ethical frameworks (Anderson et

Anderson, 2018). Davison (2000), Stahl, (2004) or

Reddy (2019) note the importance of assigning

responsibility, especially given the seriousness of the

potential consequences of an error. Some authors

suggest identifying levels of responsibility

(individual, hierarchical, collective and

organizational) or suggest assigning responsibilities

according to the role of stakeholders, according to

their decision-making (Chander, 2017; Kraemer et

al., 2011; Torresen, 2018). Others (Chander, 2017;

Kemper and Kolkman, 2019; Buhmann et al., 2020)

stresses that an organization should take

responsibility for their algorithms regardless of how

opaque they are (Malhotra et al., 2018).

2.2 AI and Bias

The algorithms that work on language are fed by

billions of data (texts, images, videos...) steeped in

our cultures. When a system becomes expert enough

to simulate a conversation and produce language that

sounds natural, it relies on the commonly accepted

ideas of the society it is addressing. Not surprisingly,

it reproduces ethically questionable historical

cultural representations. Also, for example, from the

2000s, a new and important question emerged in the

literature, under the term of social sorting (Hamet

and Michel, 2018). This trend, stemming from

surveillance studies, examines the risk that the

analysis of personal data will lead to segregation or

discrimination. Employers however no longer

hesitate to use tri- social algorithms to recruit. Banks

and insurance companies conduct scoring policies

based on these social sorting technologies. Real

estate agencies and social landlords also carry out de

facto discrimination by establishing choices of

allocation or not, using housing algorithms which

are based on the last name, on first name, on

address, on the mastery of the French language, etc.

If people tend to recruit fewer women, the algorithm

will implicitly reproduce this trend. This is a crucial

issue with AI and deep learning algorithms with

which a machine is able to learn through its own

data processing. Thus, biases in AI, based on masses

Coping with Artiﬁcial Intelligence Ethical Dilemma and Ethical Position Choices?

383

of previous data, are likely to result in discrimination

and exclusion, on a scaled-down scale by reproducing

prejudices incorporated in unprocessed data. AI,

combined with the processing of massive data, even

induces an autonomous functioning of processing

social characteristics. It is the self-learning AI itself

that produces and reproduces “social sorting”. In this

sense, AI did not invent discrimination, but they

participate in this movement, by reproducing it, even

by intensifying it.

2.3 AI and Transparency

Due to the multiplicity of actors, stakeholders and

links, algorithms, are increasingly opaque. However,

it is necessary to open the black box of algorithms,

to offer better transparency (Guidotti et al., 2020;

Buhmann et al., 2020) and then possible traceability.

As Turilli and Floridi (2009) note, transparency is not

an “ethical principle in itself but a pro-ethical

condition for enabling or impairing other ethical

practices or principles” (p.105). This is why it is

important to distinguish between the different factors

that may hinder transparency of algorithms, identify

their cause (Diakopoulos and Koliska, 2017). This

transparency is also a request from individuals in

relation to the protection of their privacy, the

confidentiality of stored data and surveillance. This

search for transparency involves being able to clarify

and decipher multiple involved and intertwined

processes, processes such as the collection of data and

their validation, the action and decision-making

processes of the chain of actors in the realization of

algorithms (initial decision, project funders, AI

researchers, analysts, developers), decision-makers

and funders of uses (for example, for health systems

or autonomous cars) (Martin, 2019). Also, the ethical

problem of the transparency of algorithms can lead us

to broader ethical questions. For example, as early as

the 1960s, some worried that advances in science and

technology could threaten the functioning of

democracy, because only a few experts were able to

truly understand complex technologies (Habermas,

1970). The technocracy hypothesis arose through

portraying a future society where experts would make

decisions based on their own value system, offering

their best solution. Then, after the experts, the AI

could also lead to a form of algorithmic

governmentality based on the statistical processing of

data communicated (voluntarily or without their

knowledge) by citizens, in particular thanks to

sensors connected, leading to a pre-elaboration of

the collective decisions thus elaborated by the

algorithms.

3 ABOUT TWO MAJORS

ETHICAL CURRENTS

MIT has initiated the Moral Machine Project, which

leverages the collective wisdom of crowds to devise

solutions to ethical dilemmas related to autonomous

vehicles controlled by AI that could potentially

cause harm to pedestrians and/or passengers in case

of malfunction. While the wisdom of the crowd is

used as an ethical reference in this project, one could

alternatively appeal to the categorical imperative, the

concept of virtue, the consequences of actions, or

other norms. Therefore, we believe it is essential to

present the major ethical theories at hand to properly

address ethical issues (Hamet and Michel, 2018).

Hence, we will discuss two primary ethical

movements, namely Kant's deontological ethics and

Hegel's consequentialist ethics. While we

acknowledge that this is not an exhaustive list, these

two ethical currents can demonstrate the significant

challenges encountered in addressing the ethical

problems posed by algorithms.

3.1 Kant’s Deontological Ethics

During the 18th century, Kant, in his work Critique

of Practical Reason (1788), attempted to answer the

question, "What shall I do?" This is Kant's second

question (after "what can I know," which is dealt

with in the Critique of Pure Reason), and the third is

"what can I hope for," which is dealt with in the

Critique of the Faculty of Judgement. At the

beginning of the Critique of Practical Reason, Kant

asks whether it is possible to construct a moral

rationalism, a supreme principle of rationality which

would be the moral law. He transforms the question

"what shall I do" into "what are the supreme

principles of morality." In this work, he promotes

the autonomy of the will. It is no longer up to the

human will to align itself with the good as an

external standard, but rather it is up to the will to

define the good as that which is universally

desirable. Kant sets out to find "practical laws," i.e.,

"objective principles valid for the will of every

reasonable being." To be objective and universal,

moral obligation must be expressed by a formal

principle, an a priori, universal, and necessary

criterion. It is therefore the autonomy of the will that

must constitute the sole principle of all moral laws

and duties. The fundamental law of pure practical

reason can be stated as follows: "Act only according

to the maxim that you can will at the same time that

it becomes a universal law."

ICEIS 2024 - 26th International Conference on Enterprise Information Systems

384

3.2 Hegel’s Ethics

Unlike Kant, Hegel relies on the experience of

human beings and aims to understand what makes

morality, rather than looking for where it should

come from.

3.2.1 Hegel’s Critique of the Kant’s Moral

Law

In The Principles of Legal Philosophy, in its second

part, Hegel addresses several criticisms of Kant’s

ethical principles and his moral law. In short:

First, the moral law is dolorist: it always

presupposes suffering in its execution, in the sense

that one must do with aversion what duty dictates.

According to Hegel, one can do one’s duty with

pleasure.

Second, circumstantial context is not taken into

account. This means that moral duties are

conditioned by the situation. Duties are not

immediate; they require reflection. Most norms

certainly work as a habit, which frees the mind,

allowing one to think about when it is necessary to

take the situation into account. There is a hierarchy

of norms. In the famous example of lying, for Hegel,

lying is seen as the least bad solution in a given

context.

Third, the moral law presents a theory that is

ahistorical. This is the most problematic point for

Hegel: Kant does not take into account the evolution

of norms and society. Hegel would say that morality

has something provisional, that it evolves over time

(referring to Descartes and his provisional morality).

So, Hegel proposes a theory of action and moral

imputation: what makes an individual’s morality is

their actions. What makes the morality of an action

is norms and morality. What makes it possible to

impute the action to the individual is ethics.

3.2.2 Hegel’s Concept of Ethical Life

In the third and final part of The Principles of the

Philosophy of Law, Hegel develops his concept of

ethical life, which encompasses the family, civil

society, the state, and world history (§142 to §360).

Ethical life is the set of norms that emanate from the

institutions that regulate the life of people: the

family, the corporations, and the state, which is the

basic institution and the condition of possibility for

other institutions. In ethical life, there are

obligations, which today are called norms.

4 AI ETHICAL ISSUES AND

ETHICAL POSITIONS

CHOICES

4.1 Lack of Ethical Positioning in the

Literature

Recent literature on these ethical problems has

proposed some interesting avenues for reflection and

action. For example, on the issue of fairness, some

authors recommend to develop a sociotechnical

framework to address and improve the fairness of

algorithms (Edwards and Veale, 2017; Selbst et al.,

2019; Wong, 2019; Abebe et al., 2020). Concerning

the issue of responsibility, for example, Shah (2017)

analysis points out that the risk of some stakeholders

failing to meet their responsibilities can be

addressed, for example, by creating separate bodies

for the ethical oversight of algorithms. However,

others show that expecting a single oversight body,

such as a research ethics committee or institutional

review board, to be 'solely responsible for ensuring

ethical rigor, utility and probity is unrealistic

(Lipworth et al., 2017). Concerning the issue of

transparency, for example, Gebru et al. (2020)

propose that the transparency constraints posed by

AI can be resolved, in part, by using standard

documentation procedures similar to those deployed

in the electronics industry. In addition, another

recent approach is the use of technical tools to test

and audit AI and decision making. This involves

checking algorithms for negative trends, such as

unfair discrimination, and auditing a prediction or

decision track in detail, (Weller, 2019; Malhotra et

al., 2018; Brundage et al. 2020). We do not question

these courses of action, but we insist that in order to

choose one of the proposed courses of action, it is

first necessary to be part of an ethical current, as the

answers differ, as we will show in the following.

4.2 Different Answers to Ethical

Questions Depending on the

Positions Chosen

AI are loaded with values (Brey et Søraker, 2009;

Kraemer et al., 2011; Mittelstadt et al., 2016;

Tsamados et al, 2021) contradicting the myth of the

long-lived neutral algorithm. Indeed, any algorithm

involves a multitude of decisions, whether

classification, prioritization, display, filtering,

learning. However, the choice of filtering

techniques, classified data, the options chosen may

well reflect a certain understanding of the world. AI

Coping with Artiﬁcial Intelligence Ethical Dilemma and Ethical Position Choices?

385

are nothing more than ideas, opinions formalized in

code, and in no way escape the subjectivity of

developers, managers, contractors or society.

Algorithmic coding contains a wide spectrum of

standards that can range from moral injunctions to

unconsciously integrated norms. We want to show

that choosing these solutions without first choosing

an ethical position is a bad approach. Indeed, the

solution chosen can only depend on the ethical

position initially taken. Responses to these ethical

problems will differ depending on whether one

adheres to a deontological Kantian ethics or a

consequentialist Hegelian ethics.

With regards to the ethical theme of biases,

according to Kantian’s ethics, AI in its entirety must

respond to a categorical imperative of non-

discrimination and justice. AI must not discriminate

or reproduce discriminations. This could be one of

the first categorical imperatives assigned to this

ethical theme. It would then be a question of

knowing how to put in place this ethic approach to

duty. Consequentialist’s, on the other hand, will be

interested in the consequences of these biases and

discriminations with regards to the well-being of

society. Some discrimination can be accepted. They

may consider that discrimination against a minority

is not detrimental to social well-being and therefore

that it is acceptable.

With regard to the theme of responsibility, we

find the two positions defined above. For Kantians

Ethics, faced with the tram dilemma type (thought

experiment which offers a person a choice of action,

knowing that if he acts, his gesture will benefit a

group of people, but will harm a person), dilemma

which is very similar to the dilemmas posed to the

autonomous car, it is not possible that a person will

be killed because of my action. For consequentialists

ethics, faced with this dilemma, between one or

more people, between young or old, depending on

the social utility of the person, we can make a

choice. Choosing to sacrifice one person to save five

can be understood according to this morality.

Finally, regarding the theme of transparency, the

ethical imperative leads us to consider the initiator

of the standard - whether it's the machine or the

human. Is it ethical to delegate decision-making to a

machine? Can a machine have good intentions?

According to Hegelian ethics, we must identify the

consequences of a loss of control and human

mastery. The time, financial, and fatigue savings

generated by delegating decision-making to a

machine can justify certain costs, such as the loss of

certain degrees of freedom and generalized

surveillance. On the other hand, Kantian ethics

emphasizes individual reflection. In order to create

contextualized ethical norms, Hegel proposes three

levels of action: family, civil society, and state. We

can summarize our approach in the following table1:

Table 1.

AI Ethical questions -

Synthetic formulation

Questions raised according

to ethical theorical positions

Bias

Questioning the

integration of societal

values in AI. The rules

included in the

algorithms are not

neutral and convey

conscious and

unconscious values of

the developers, the

organizations, the

companies.

Kantian ethics: questioning

of Universal moral rules to be

integrated into algorithms

(absence of discrimination)

and to be imposed on society.

Fairness couldn’t’ be

accepted, never.

Hegel ethics and

consequentialists current:

questioning the consequences

of the biases generated. Some

on theses could be accepted.

Responsibility

Amoral AI;

Identification of

responsibilities? Dilution

of responsibilities;

Revealing ethics and the

system as an actor in its

own right

Kantian ethics: questioning

a “categorical imperative”

(the autonomous car must not

kill, for example).

Responsibility must always

be accurately attributed

Hegel ethics and

consequentialists current:

questioning the consequences

of ethical dilemmas. A trade-

off between different

responsibilities is possible

Transparency

Transparency of the

process: data/algorithmic

processing/ effects

Traceability

Kantian ethics:

questioning the initiator of

the categorical imperative

(Human or algorithm)

Hegel ethics and

Consequentialists current:

questioning the

consequences

of the

excesses of the loss of

human control

5 CONCLUSIONS, DISCUSSIONS

AND PERSPECTIVES

In summary, in the proliferation of ethical problems

posed by AI, three paths of ethical questioning

related to AI have been identified: the question

of algorithmic biases, the question of responsibility,

ICEIS 2024 - 26th International Conference on Enterprise Information Systems

386

the question of transparency. We show that

these themes need to be according to the ethical

theory in which they are questioned, (Deontological,

Consequentialists).

The contributions of this article are therefore

twofold. First, we have consolidated the literature on

the topic of the ethics of AI, by putting forward a

synthesis of the main ethical questions. This

synthesis sheds light on this confused field. We have

also shown, with support of the main ethical

theoretical currents, that these questions can lead to

very different answers. Thus, in terms of practical

ethics, it is not enough for organizations to identify

ethical issues and propose ethical charters. The first

task, which underlies the whole, is indeed to choose

an ethical theoretical position.

REFERENCES

Abebe R., Barocas S., Kleinberg J., Levy K., Raghavan

M., Robinson D.G., (2020) Roles for computing in

social change, in Proceedings of the 2020 Conference

on Fairness, Accountability, and Transparency, pp.

252–260.

Aggarwal N., (2021), The norms of algorithmic credit

scoring, The Cambridge Law Journal, vol. 80, n°1,

pp. 42–73.

Anderson, M., & Anderson, S. L. (2018). GenEth: A

general ethical dilemma analyzer. Paladyn, Journal of

Behavioral Robotics, 9(1), 337-357.

Berreby F., Bourgne G., Ganasc J.G., (2017) A declarative

modular framework for representing and applying

ethical principles, In 16th Conference on Autonomous

Agents and Multiagent Systems, 2017, May.

Brey P., Søraker J.H.,(2009) Philosophy of computing and

information technology. In Philosophy of technology

and engineering sciences, pp. 1341–1407, North-

Holland.

Brundage M., Avin S., Wang J., Belfield H., Krueger G.,

Hadfield G., Khlaaf H. et al., (2020) Toward

trustworthy AI development: mechanisms for

supporting verifiable claims, arXiv preprint

arXiv:2004.07213.

Buhmann, A., Paßmann J., Fieseler J.C., (2020) Managing

algorithmic accountability: Balancing reputational

concerns, engagement strategies, and the potential of

rational discourse, Journal of Business Ethics, vol.

163, n°2, pp. 265–280.

Chander A., (2017) The racist algorithm?, Michigan Law

Review, vol. 115, n° 6, pp. 1023–1045,.

Chatterjee S., Sarker S., Fuller M., (2009) Ethical

information systems development: A Baumanian

postmodernist perspective, Journal of the Association

for Information Systems, vol. 10, n°11, pp. 787–815.

Davison R.M., (2000) Professional ethics in information

systems: A personal perspective, Communications of

the association for Information Systems, vol.3,n°1.

De George R.T. (1999), Business ethics and the

information age, Business and Society Review, vol

104, n°3, pp. 261–278.

Diakopoulos N., Koliska M., (2017), Algorithmic

transparency in the news media, Digit Journal, vol 5,

n°7, pp.809–828.

Edwards L., Veale M., (2017) Slave to the algorithm?

Why a right to explanation is probably not the remedy

you are looking for, SSRN Electron J.,

Gebru T., Morgenstern J., Vecchione B., Vaughan J.W.,

Wallach H., Lii H.D., Crawford H,K., (2020)

Datasheets for datasets, Communications of the ACM,

vol. 64, n° 12, pp. 86–92, pp.

Guidotti R., Monreale A., Ruggieri S., Turini F., Giannotti

F., Pedreschi D., (2018) A survey of methods for

explaining black box models, ACM computing surveys

(CSUR), vol 51, n°5, pp. 1– 42, 2018.

Habermas J. (1970), Towards a theory of communicative

competence, Inquiry, vol. 13, n°1–4, pp. 360–375.

Hamet J., Michel S., (2018), Les questionnements éthiques

en systèmes d’information, Revue française de

gestion, vol. 44, n°271, pp. 99–129.

Kant E., (1905 ,orig 1787) Critique de la raison pratique,

Editions Flammarion.

Karppi T., (2018), The computer said so: on the ethics,

effectiveness, and cultural techniques of predictive

policing, Soc MedAI Soc, vol 4, n° 2,.

Kemper J., Kolkman D.,( 2019), Transparent to whom?

No algorithmic accountability without a critical

audience, Information, Communication & Society, vol

22, n°14, pp. 2081–2096.

Kraemer F., Van Overveld K., Peterson M., (2011), Is

there an ethics of algorithms?, Ethics and information

technology, vol.13, n°3, pp.251–260.

Light B., McGrath K., (2010 ), Ethics and social

networking sites: a disclosive analysis of Facebook,

Information Technology & People.

Lipworth W., Mason P.H., Kerridge I., Ioannidis J.P.A.,

(2017), Ethics and epistemology in big data research, J

Bioethical Inq, vol. 14, n°4, pp.489– 500.

Malhotra C., Kotwal V., Dalal S., (2018), Ethical

framework for machine learning, ITU Kaleidoscope:

machine learning for a 5G Future, pp. 1–8. Santa Fe:

IEEE,.

Martin K., (2019), Designing ethical algorithms, MIS

Quarterly Executive June, 2019.

Milano S., Taddeo M., Floridi L., Recommender systems

and their ethical challenges, Ai & Society, vol.35, n°4,

pp.957–967, 2020.

Mittelstadt B.D., Allo P., Taddeo M., Wachter S., Floridi

L., The ethics of algorithms: Mapping the debate. Big.

Data & Society, vol.3, n°2; 2016.

Neyland D., (2016), Bearing accountable witness to the

ethical algorithmic system, Science, Technology &

Human Values, vol.41, n°1, pp.50–76.

Obermeyer Z., Powers B., Vogeli C., Mullainathan S.,

(2019), Dissecting racial bias in an algorithm used to

manage the health of populations, Science,vol.366,

n°6464, pp. 447–453,.

Coping with Artiﬁcial Intelligence Ethical Dilemma and Ethical Position Choices?

387

Reddy E., Cakici B., Ballestero A., (2019), Beyond

mystery: putting algorithmic accountability in context,

Big Data Soc, vol.6, n°1.

Rubel A., Castro C., Pham A., (2019), Agency laundering

and information technologies, Ethical Theory Moral

Pract, vol. 22, n°4, pp.1017–1041.

Selbst A.D., Boyd D., Friedler S.A., Venkatasubramanan

S., Vertesi J., (2019), Fairness and abstraction in

sociotechnical systems, Proceedings of the Conference

on Fairness, Accountability, and Transparency, pp.

59–68. Atlanta, GA, USA: ACM Press,

Shah H., (2018), Algorithmic accountability, Philos Trans

R Soc A: Math Phys Eng Sci, Vol 376.

Seaver N., (2017), Algorithms as culture: Some tactics for

the ethnography of AI, Big data & society, vol. 4, n°2

Stahl B.C., (2004), Information, ethics, and computers:

The problem of autonomous moral agents, Minds

Mach, vol. 14, n°1, pp. 67–83, 2004.

Torresen J., (2018), A review of future and ethical

perspectives of robotics and AI, Frontiers in Robotics

and AI, vol. 4, pp. 75.

Tsamados, A. , Aggarwal, N. , Cowls, J. and Morley J.,

and Roberts H., Taddeo M., Floridi, L., (2021), The

Ethics of Algorithms: Key Problems and Solutions 37

AI & Society 215

Floridi L., (2021), The ethics of algorithms: key problems

and solutions, AI & SOCIETY, pp. 1–16.

Turilli M., Floridi L., (2009), The ethics of information

transparency, Ethics and Information Technology,

vol.11, n°2, pp.105–112

Weller, A. (2019), Transparency: motivations and

challenges. In Explainable AI: interpreting, explaining

and visualizing deep learning (pp. 23-40). Cham:

Springer International Publishing.

Wieringa, M., (2020), What to account for when

accounting for algorithms: a systematic literature

review on algorithmic accountability, In Proceedings

of the 2020 conference on fairness, accountability, and

transparency, pp. 1-18, January 2020.

Wong P.H., (2019), Democratizing algorithmic fairness.

Philos Technol,.

Yu, H., Shen, Z., Miao, C., Leung, C., Lesser, V. R., &

Yang, Q. (2018). Building ethics into artificial

intelligence. arXiv preprint arXiv:1812.02953.

ICEIS 2024 - 26th International Conference on Enterprise Information Systems

388