Theoretical research

Klenk, M. (2022). AI Design and Governance. The State of AI Ethics Report 6, Montreal AI Ethics Institute, 150-152.

"Another new addition to this report which builds on our push towards moving from principles to practice is the chapter on AI Design and Governance which has the goal of dissecting the entire ecosystem around AI and the AI lifecycle itself to gain a very deep understanding of the choices and decisions that lead to some of the ethical issues that arise in AI. It constitutes about one-sixth of the report and is definitely something that I would encourage you to read in its entirety to gain some new perspectives on how we can actualize Responsible AI."

Van den Hoven, J., Vermaas, P., & van de Poel, I. (2015). Handbook of Ethics, Values, and Technological Design. Springer, Netherlands.

This handbook enumerates every aspect of incorporating moral and societal values into technology design, reflects the fact that the latter has moved on from strict functionality to become sensitive to moral and social values such as sustainability and accountability. Aimed at a broad readership that includes ethicists, policy makers and designers themselves, it proffers a detailed survey of how technological, and institutional, design must now reflect awareness of ethical factors such as sustainability, human well-being, privacy, democracy and justice, inclusivity, trust, accountability, and responsibility (both social and environmental). Edited by a trio of highly experienced academic philosophers with a specialized interest in the ethical dimensions of technology and human creativity, this syncretic handbook collates an array of published material and offers a studied, practical introduction to the field. The volume addresses myriad aspects at the intersection of technology design and ethics, enabling designers to adopt a constructive approach in anticipating, preventing, and resolving societal and ethical issues affecting their work. It covers underlying theory; discrete values such as democracy, human well-being, sustainability and justice; and application domains themselves, which include architecture, bio- and nanotechnology, and military hardware. As the first exhaustive survey of a field whose importance is characterized by almost exponential growth, it represents a compelling addition to a formerly atomized literature.​

Klenk, M. (2022). (Online) manipulation: sometimes hidden, always careless. Review of Social Economy 80, 2: 85-105.

Ever-increasing numbers of human interactions with intelligent software agents, online and offline, and their increasing ability to influence humans have prompted a surge in attention toward the concept of (online) manipulation. Several scholars have argued that manipulative influence is always hidden. But manipulation is sometimes overt, and when this is acknowledged the distinction between manipulation and other forms of social influence becomes problematic. Therefore, we need a better conceptualisation of manipulation that allows it to be overt and yet clearly distinct from related concepts of social influence. I argue that manipulation is careless influence, show how this account helps to alleviate the shortcomings of the hidden influence view of manipulation, and derive implications for digital ethics.

Marin, L. (2022). Enactive Principles for the Ethics of User Interactions on Social Media: How to Overcome Systematic Misunderstandings Through Shared Meaning-Making. Topoi, 1-13.

This paper proposes three principles for the ethical design of online social environments aiming to minimise the unintended harms caused by users while interacting online, specifically by enhancing the users’ awareness of the moral load of their interactions. Such principles would need to account for the strong mediation of the digital environment and the particular nature of user interactions: disembodied, asynchronous, and ambiguous intent about the target audience. I argue that, by contrast to face to face interactions, additional factors make it more difficult for users to exercise moral sensitivity in an online environment. An ethics for social media user interactions is ultimately an ethics of human relations mediated by a particular environment; hence I look towards an enactive inspired ethics in formulating principles for human interactions online to enhance or at least do not hinder a user’s moral sensitivity. This enactive take on social media ethics supplements classical moral frameworks by asking us to focus on the relations established through the interactions and the environment created by those interactions.

Santoni de Sio, F. & Van den Hoven, J. (2018). Meaningful Human Control over Autonomous Systems: A Philosophical Account. Frontiers in Robotics and AI, 5:15.

Debates on lethal autonomous weapon systems have proliferated in the past 5 years. Ethical concerns have been voiced about a possible raise in the number of wrongs and crimes in military operations and about the creation of a “responsibility gap” for harms caused by these systems. To address these concerns, the principle of “meaningful human control” has been introduced in the legal–political debate; according to this principle, humans not computers and their algorithms should ultimately remain in control of, and thus morally responsible for, relevant decisions about (lethal) military operations. However, policy-makers and technical designers lack a detailed theory of what “meaningful human control” exactly means. In this paper, we lay the foundation of a philosophical account of meaningful human control, based on the concept of “guidance control” as elaborated in the philosophical debate on free will and moral responsibility. Following the ideals of “Responsible Innovation” and “Value-sensitive Design,” our account of meaningful human control is cast in the form of design requirements. We identify two general necessary conditions to be satisfied for an autonomous system to remain under meaningful human control: first, a “tracking” condition, according to which the system should be able to respond to both the relevant moral reasons of the humans designing and deploying the system and the relevant facts in the environment in which the system operates; second, a “tracing” condition, according to which the system should be designed in such a way as to grant the possibility to always trace back the outcome of its operations to at least one human along the chain of design and operation. As we think that meaningful human control can be one of the central notions in ethics of robotics and AI, in the last part of the paper, we start exploring the implications of our account for the design and use of non-military autonomous systems, for instance, self-driving cars.

Santoni de Sio, F., & Mecacci, G. (2021). Four Responsibility Gaps with Artificial Intelligence: Why They Matter and How to Address them. Philosophy & Technology 34, 1057-1084.

The notion of “responsibility gap” with artificial intelligence (AI) was originally introduced in the philosophical debate to indicate the concern that “learning automata” may make more difficult or impossible to attribute moral culpability to persons for untoward events. Building on literature in moral and legal philosophy, and ethics of technology, the paper proposes a broader and more comprehensive analysis of the responsibility gap. The responsibility gap, it is argued, is not one problem but a set of at least four interconnected problems – gaps in culpability, moral and public accountability, active responsibility—caused by different sources, some technical, other organisational, legal, ethical, and societal. Responsibility gaps may also happen with non-learning systems. The paper clarifies which aspect of AI may cause which gap in which form of responsibility, and why each of these gaps matter. It proposes a critical review of partial and non-satisfactory attempts to address the responsibility gap: those which present it as a new and intractable problem (“fatalism”), those which dismiss it as a false problem (“deflationism”), and those which reduce it to only one of its dimensions or sources and/or present it as a problem that can be solved by simply introducing new technical and/or legal tools (“solutionism”). The paper also outlines a more comprehensive approach to address the responsibility gaps with AI in their entirety, based on the idea of designing socio-technical systems for “meaningful human control", that is systems aligned with the relevant human reasons and capacities.

Mollen, J. (under review). Moving out of the Human Vivarium: Live-in Laboratories and the Right to Withdraw.

Durán, J. (2021). Dissecting Scientific Explanation in AI (sXAI): A Case for Medicine and Healthcare. Artificial Intelligence 297, 103498.

Explanatory AI (XAI) is on the rise, gaining enormous traction with the computational community, policymakers, and philosophers alike. This article contributes to this debate by first distinguishing scientific XAI (sXAI) from other forms of XAI. It further advances the structure for bona fide sXAI, while remaining neutral regarding preferences for theories of explanations. Three core components are under study, namely, i) the structure for bona fide sXAI, consisting in elucidating the explanans, the explanandum, and the explanatory relation for sXAI: ii) the pragmatics of explanation, which includes a discussion of the role of multi-agents receiving an explanation and the context within which the explanation is given; and iii) a discussion on Meaningful Human Explanation, an umbrella concept for different metrics required for measuring the explanatory power of explanations and the involvement of human agents in sXAI. The kind of AI systems of interest in this article are those utilized in medicine and the healthcare system. The article also critically addresses current philosophical and computational approaches to XAI. Amongst the main objections, it argues that there has been a long-standing interpretation of classifications as explanation, when these should be kept separate.

Durán, J. & Jongsma, K. (2021). Who Is Afraid of Black-Box Algorithms? On the Epistemological and Ethical Basis of Trust in Medical AI. Journal of Medical Ethics 47, 329-335

The use of black box algorithms in medicine has raised scholarly concerns due to their opaqueness and lack of trustworthiness. Concerns about potential bias, accountability and responsibility, patient autonomy and compromised trust transpire with black box algorithms. These worries connect epistemic concerns with normative issues. In this paper, we outline that black box algorithms are less problematic for epistemic reasons than many scholars seem to believe. By outlining that more transparency in algorithms is not always necessary, and by explaining that computational processes are indeed methodologically opaque to humans, we argue that the reliability of algorithms provides reasons for trusting the outcomes of medical artificial intelligence (AI). To this end, we explain how computational reliabilism, which does not require transparency and supports the reliability of algorithms, justifies the belief that results of medical AI are to be trusted. We also argue that several ethical concerns remain with black box algorithms, even when the results are trustworthy. Having justified knowledge from reliable indicators is, therefore, necessary but not sufficient for normatively justifying physicians to act. This means that deliberation about the results of reliable algorithms is required to find out what is a desirable action. Thus understood, we argue that such challenges should not dismiss the use of black box algorithms altogether but should inform the way in which these algorithms are designed and implemented. When physicians are trained to acquire the necessary skills and expertise, and collaborate with medical informatics and data scientists, black box algorithms can contribute to improving medical care.

Marin, L. (2021). Sharing (Mis) information on Social Networking Sites. An Exploration of the norms for Distributing Content Authored by Others. Ethics and Information Technology, 23(3), 363-372.

This article explores the norms that govern regular users’ acts of sharing content on social networking sites. Many debates on how to counteract misinformation on Social Networking Sites focus on the epistemic norms of testimony, implicitly assuming that the users’ acts of sharing should fall under the same norms as those for posting original content. I challenge this assumption by proposing a non-epistemic interpretation of (mis) information sharing on social networking sites which I construe as infrastructures for forms of life found online. Misinformation sharing belongs more in the realm of rumour spreading and gossiping rather than in the information-giving language games. However, the norms for sharing cannot be fixed in advance, as these emerge at the interaction between the platforms’ explicit rules, local norms established by user practices, and a meta-norm of sociality. This unpredictability does not leave us with a normative void as an important user responsibility still remains, namely that of making the context of the sharing gesture explicit. If users will clarify how their gestures of sharing are meant to be interpreted by others, they will implicitly assume responsibility for possible misunderstandings based on omissions, and the harms of shared misinformation can be diminished.

Pozzi, G. and Durán, Juan M. (under review). Informativeness and Epistemic Injustice in Explanatory Medical Machine Learning.

Pozzi, G. (under review). Automated Opioid Risk Scores: A Case for Machine Learning-Induced Epistemic Injustice in Healthcare.