We propose a formal framework for interpretable machine learning. Combining elements from statistical learning, causal interventionism, and decision theory, we design an idealised explanation game in which players collaborate to find the best explanation for a given algorithmic prediction. Through an iterative procedure of questions and answers, the players establish a three-dimensional Pareto frontier that describes the optimal trade-offs between explanatory accuracy, simplicity, and relevance. Multiple rounds are played at different levels of abstraction, allowing the players to explore overlapping causal (...) patterns of variable granularity and scope. We characterise the conditions under which such a game is almost surely guaranteed to converge on a optimal explanation surface in polynomial time, and highlight obstacles that will tend to prevent the players from advancing beyond certain explanatory thresholds. The game serves a descriptive and a normative function, establishing a conceptual space in which to analyse and compare existing proposals, as well as design new and improved solutions. (shrink)

We propose a formal framework for interpretable machine learning. Combining elements from statistical learning, causal interventionism, and decision theory, we design an idealisedexplanation gamein which players collaborate to find the best explanation(s) for a given algorithmic prediction. Through an iterative procedure of questions and answers, the players establish a three-dimensional Pareto frontier that describes the optimal trade-offs between explanatory accuracy, simplicity, and relevance. Multiple rounds are played at different levels of abstraction, allowing the players to explore overlapping causal patterns of (...) variable granularity and scope. We characterise the conditions under which such a game is almost surely guaranteed to converge on a (conditionally) optimal explanation surface in polynomial time, and highlight obstacles that will tend to prevent the players from advancing beyond certain explanatory thresholds. The game serves a descriptive and a normative function, establishing a conceptual space in which to analyse and compare existing proposals, as well as design new and improved solutions. (shrink)

Some recent developments in Artificial Intelligence—especially the use of machine learning systems, trained on big data sets and deployed in socially significant and ethically weighty contexts—have led to a number of calls for “transparency”. This paper explores the epistemological and ethical dimensions of that concept, as well as surveying and taxonomising the variety of ways in which it has been invoked in recent discussions. Whilst “outward” forms of transparency may be straightforwardly achieved, what I call “functional” transparency about the inner (...) workings of a system is, in many cases, much harder to attain. In those situations, I argue that contestability may be a possible, acceptable, and useful alternative so that even if we cannot understand how a system came up with a particular output, we at least have the means to challenge it. (shrink)

The literature on ethics and user attitudes towards AVs discusses user concerns in relation to automation; however, we show that there are additional relevant issues at stake. To assess adolescents’ attitudes regarding the ‘car of the future’ as presented by car manufacturers, we conducted two studies with over 400 participants altogether. We used a mixed methods approach in which we combined qualitative and quantitative methods. In the first study, our respondents appeared to be more concerned about other aspects of AVs (...) than automation. Instead, their most commonly raised concerns were the extensive use of AI, recommender systems, and related issues of autonomy, invasiveness and personal privacy. The second study confirmed that several AV impacts were negatively perceived. The responses were, however, ambivalent. This confirms previous research on AV attitudes. On one hand, the AV features were perceived as useful, while on the other hand, their impacts were negatively assessed. We followed theoretical insights form futures studies and responsible research and innovation, which helped to identify that there are additional user concerns than what has been previously discussed in the literature on public attitudes and ethics of AVs, as well what has been envisioned by car manufactures. (shrink)

Ethical dissonance arises from conflicts between beliefs or behaviors and affects ethical factors such as normality or conformity. This paper proposes a weak signal-oriented framework to investigate ethical dissonance from experiences linked to human–machine interactions. It is based on a systems engineering principle called human-systems inclusion, which considers any experience feedback of weak signals as beneficial to learn. The framework studies weak signal-based scenarios from testimonies of individual experiences and these scenarios are assessed by other people. For this purpose, the (...) framework proposes several databases as sources of weak signals, formalization tools of experience feedback of weak signals, models of references of conformity, ethical factors, and a list of examples of ethical dissonance. It also includes sequential steps to make the latter credible regarding the results from the experimental protocols. The framework was used to investigate ethical dissonance by analyzing experiences pertaining to achieving inclusive mobility. The first example focuses on ethical dissonance in terms of hindrance that goes against autonomous mobility due to a misunderstanding of how the system functions in terms of negative and positive emotions. Two other examples present possible ethical dissonance when the use of safety systems such as car driver-assistance systems, may create danger. Investigating ethical dissonance can then help system-inclusive design or evaluation processes by taking into account scenarios from weak signal-based experiences and making them credible. (shrink)

Definition of the problem The umbrella term “explicability” refers to the reduction of opacity of artificial intelligence (AI) systems. These efforts are challenging for medical AI applications because higher accuracy often comes at the cost of increased opacity. This entails ethical tensions because physicians and patients desire to trace how results are produced without compromising the performance of AI systems. The centrality of explicability within the informed consent process for medical AI systems compels an ethical reflection on the trade-offs. Which (...) levels of explicability are needed to obtain informed consent when utilizing medical AI? Arguments We proceed in five steps: First, we map the terms commonly associated with explicability as described in the ethics and computer science literature, i.e., disclosure, intelligibility, interpretability, and explainability. Second, we conduct a conceptual analysis of the ethical requirements for explicability when it comes to informed consent. Third, we distinguish hurdles for explicability in terms of epistemic and explanatory opacity. Fourth, this then allows to conclude the level of explicability physicians must reach and what patients can expect. In a final step, we show how the identified levels of explicability can technically be met from the perspective of computer science. Throughout our work, we take diagnostic AI systems in radiology as an example. Conclusion We determined four levels of explicability that need to be distinguished for ethically defensible informed consent processes and showed how developers of medical AI can technically meet these requirements. (shrink)

Recent years have witnessed intensive efforts to specify which requirements ethical artificial intelligence (AI) must meet. General guidelines for ethical AI consider a varying number of principles important. A frequent novel element in these guidelines, that we have bundled together under the term explicability, aims to reduce the black-box character of machine learning algorithms. The centrality of this element invites reflection on the conceptual relation between explicability and the four bioethical principles. This is important because the application of general ethical (...) frameworks to clinical decision-making entails conceptual questions: Is explicability a free-standing principle? Is it already covered by the well-established four bioethical principles? Or is it an independent value that needs to be recognized as such in medical practice? We discuss these questions in a conceptual-ethical analysis, which builds upon the findings of an empirical document analysis. On the example of the medical specialty of radiology, we analyze the position of radiological associations on the ethical use of medical AI. We address three questions: Are there references to explicability or a similar concept? What are the reasons for such inclusion? Which ethical concepts are referred to? (shrink)

Most engineers Fwork within social structures governing and governed by a set of values that primarily emphasise economic concerns. The majority of innovations derive from these loci. Given the effects of these innovations on various communities, it is imperative that the values they embody are aligned with those societies. Like other transformative technologies, artificial intelligence systems can be designed by a single organisation but be diffused globally, demonstrating impacts over time. This paper argues that in order to design for this (...) broad stakeholder group, engineers must adopt a systems thinking approach that allows them to understand the sociotechnicity of artificial intelligence systems across sociocultural domains. It claims that value sensitive design, and envisioning cards in particular, provides a solid first step towards helping designers harmonise human values, understood across spatiotemporal boundaries, with economic values, rather than the former coming at the opportunity cost of the latter. (shrink)

Healthcare is becoming increasingly automated with the development and deployment of care robots. There are many benefits to care robots but they also pose many challenging ethical issues. This paper takes care robots for the elderly as the subject of analysis, building on previous literature in the domain of the ethics and design of care robots. Using the value sensitive design approach to technology design, this paper extends its application to care robots by integrating the values of care, values that (...) are specific to AI, and higher-scale values such as the United Nations Sustainable Development Goals. The ethical issues specific to care robots for the elderly are discussed at length alongside examples of specific design requirements that work to ameliorate these ethical concerns. (shrink)

Research on the ethics of algorithms has grown substantially over the past decade. Alongside the exponential development and application of machine learning algorithms, new ethical problems and solutions relating to their ubiquitous use in society have been proposed. This article builds on a review of the ethics of algorithms published in 2016, 2016). The goals are to contribute to the debate on the identification and analysis of the ethical implications of algorithms, to provide an updated analysis of epistemic and normative (...) concerns, and to offer actionable guidance for the governance of the design, development and deployment of algorithms. (shrink)

Research on the ethics of algorithms has grown substantially over the past decade. Alongside the exponential development and application of machine learning algorithms, new ethical problems and solutions relating to their ubiquitous use in society have been proposed. This article builds on a review of the ethics of algorithms published in 2016, 2016). The goals are to contribute to the debate on the identification and analysis of the ethical implications of algorithms, to provide an updated analysis of epistemic and normative (...) concerns, and to offer actionable guidance for the governance of the design, development and deployment of algorithms. (shrink)

With recent advancements in systems engineering and artificial intelligence, autonomous agents are increasingly being called upon to execute tasks that have normative relevance. These are tasks that directly—and potentially adversely—affect human well-being and demand of the agent a degree of normative-sensitivity and -compliance. Such norms and normative principles are typically of a social, legal, ethical, empathetic, or cultural nature. Whereas norms of this type are often framed in the abstract, or as high-level principles, addressing normative concerns in concrete applications of (...) autonomous agents requires the refinement of normative principles into explicitly formulated practical rules. This paper develops a process for deriving specification rules from a set of high-level norms, thereby bridging the gap between normative principles and operational practice. This enables autonomous agents to select and execute the most normatively favourable action in the intended context premised on a range of underlying relevant normative principles. In the translation and reduction of normative principles to SLEEC rules, we present an iterative process that uncovers normative principles, addresses SLEEC concerns, identifies and resolves SLEEC conflicts, and generates both preliminary and complex normatively-relevant rules, thereby guiding the development of autonomous agents and better positioning them as normatively SLEEC-sensitive or SLEEC-compliant. (shrink)

With recent advancements in systems engineering and artificial intelligence, autonomous agents are increasingly being called upon to execute tasks that have normative relevance. These are tasks that directly—and potentially adversely—affect human well-being and demand of the agent a degree of normative-sensitivity and -compliance. Such norms and normative principles are typically of a social, legal, ethical, empathetic, or cultural (‘SLEEC’) nature. Whereas norms of this type are often framed in the abstract, or as high-level principles, addressing normative concerns in concrete applications (...) of autonomous agents requires the refinement of normative principles into explicitly formulated practical rules. This paper develops a process for deriving specification rules from a set of high-level norms, thereby bridging the gap between normative principles and operational practice. This enables autonomous agents to select and execute the most normatively favourable action in the intended context premised on a range of underlying relevant normative principles. In the translation and reduction of normative principles to SLEEC rules, we present an iterative process that uncovers normative principles, addresses SLEEC concerns, identifies and resolves SLEEC conflicts, and generates both preliminary and complex normatively-relevant rules, thereby guiding the development of autonomous agents and better positioning them as normatively SLEEC-sensitive or SLEEC-compliant. (shrink)

Organizations are making massive investments in artificial intelligence, and recent demonstrations and achievements highlight the immense potential for AI to improve organizational and human welfare. Yet realizing the potential of AI necessitates a better understanding of the various ethical issues involved with deciding to use AI, training and maintaining it, and allowing it to make decisions that have moral consequences. People want organizations using AI and the AI systems themselves to behave ethically, but ethical behavior means different things to different (...) people, and many ethical dilemmas require trade-offs such that no course of action is universally considered ethical. How should organizations using AI—and the AI itself—process ethical dilemmas where humans disagree on the morally right course of action? Though a variety of ethical AI frameworks have been suggested, these approaches do not adequately address how people make ethical evaluations of AI systems or how to incorporate the fundamental disagreements people have regarding what is and is not ethical behavior. Drawing on moral foundations theory, we theorize that a person will perceive an organization’s use of AI, its data procedures, and the resulting AI decisions as ethical to the extent that those decisions resonate with the person’s moral foundations. Since people hold diverse moral foundations, this highlights the crucial need to consider individual moral differences at multiple levels of AI. We discuss several unresolved issues and suggest potential approaches for thinking about conflicts in moral judgments concerning AI. (shrink)

The design of collaborative robotics, such as driver-assisted operations, engineer a potential automation of decision-making predicated on unobtrusive data gathering of human users. This form of ‘somatic surveillance’ increasingly relies on behavioural biometrics and sensory algorithms to verify the physiology of bodies in cabin interiors. Such processes secure cyber-physical space, but also register user capabilities for control that yield data as insured risk. In this technical re-formation of human–machine interactions for control and communication ‘a dissonance of attribution’ :7684, 2019. https://doi.org/10.1073/pnas.1805770115) (...) is created between perceptions of phenomena, materials and decision-making. This reconfigures relations not only between humans and machines, objects and subjects, but possibly disrupts attributive functions in the social system of Law. What it requires is shifting a legal accountability for action from a sovereignty of the human to a new materialist account based on a ‘cognitive assemblage’ between physiological data, computation and algorithmic sensing. This paper investigates the function of law as a guidance system to acknowledge this account of sensory and algorithmic computation as autonomous ‘sensing agents’ that may be accountable in situations of risk. This assemblage of robotic computation and sensory determination requires a clearer legal differentiation across the current static terminologies of person, property, liability and rights that maintain strict separations of object from subject. To neglect this, we argue, law will solely impute attributions of error to humans despite evidence of operation via mutual control. (shrink)

In the development of governmental policy for artificial intelligence that is informed by ethics, one avenue currently pursued is that of drawing on “AI Ethics Principles”. However, these AI Ethics Principles often fail to be actioned in governmental policy. This paper proposes a novel framework for the development of ‘Actionable Principles for AI’. The approach acknowledges the relevance of AI Ethics Principles and homes in on methodological elements to increase their practical implementability in policy processes. As a case study, elements (...) are extracted from the development process of the Ethics Guidelines for Trustworthy AI of the European Commission’s “High Level Expert Group on AI”. Subsequently, these elements are expanded on and evaluated in light of their ability to contribute to a prototype framework for the development of 'Actionable Principles for AI'. The paper proposes the following three propositions for the formation of such a prototype framework: preliminary landscape assessments; multi-stakeholder participation and cross-sectoral feedback; and, mechanisms to support implementation and operationalizability. (shrink)

Artificial intelligence (AI) plays a rapidly increasing role in clinical care. Many of these systems, for instance, deep learning-based applications using multilayered Artificial Neural Nets, exhibit epistemic opacity in the sense that they preclude comprehensive human understanding. In consequence, voices from industry, policymakers, and research have suggested trust as an attitude for engaging with clinical AI systems. Yet, in the philosophical and ethical literature on medical AI, the notion of trust remains fiercely debated. Trust skeptics hold that talking about trust (...) in nonhuman agents constitutes a category error and worry about the concept being misused for ethics washing. Proponents of trust have responded to these worries from various angles, disentangling different concepts and aspects of trust in AI, potentially organized in layers or dimensions. Given the substantial disagreements across these accounts of trust and the important worries about ethics washing, we embrace a diverging strategy here. Instead of aiming for a positive definition of the elements and nature of trust in AI, we proceed ex negativo, that is we look at cases where trust or distrust are misplaced. Comparing these instances with trust expedited in doctor–patient relationships, we systematize these instances and propose a taxonomy of both misplaced trust and distrust. By inverting the perspective and focusing on negative examples, we develop an account that provides useful ethical constraints for decisions in clinical as well as regulatory contexts and that highlights how we should not engage with medical AI. (shrink)

Certain characteristics make machine learning a powerful tool for processing large amounts of data, and also particularly unsuitable for explanatory purposes. There are worries that its increasing use in science may sideline the explanatory goals of research. We analyze the key characteristics of ML that might have implications for the future directions in scientific research: epistemic opacity and the ‘theory-agnostic’ modeling. These characteristics are further analyzed in a comparison of ML with the traditional statistical methods, in order to demonstrate what (...) it is specifically that makes ML methodology substantially unsuitable for reaching explanations. The analysis is given broader philosophical context by connecting it with the views on the role of prediction and explanation in science, their relationship, and the value of explanation. We proceed to show, first, that ML disrupts the relationship between prediction and explanation commonly understood as a functional relationship. Then we show that the value of explanation is not exhausted in purely predictive functions, but rather has a ubiquitously recognized value for both science and everyday life. We then invoke two hypothetical scenarios with different degrees of automatization of science, which help test our intuitions on the role of explanation in science. The main question we address is whether ML will reorient or otherwise impact our standard explanatory practice. We conclude with a prognosis that ML would diversify science into purely predictively oriented research based on ML-like techniques and, on the other hand, remaining faithful to anthropocentric research focused on the search for explanation. (shrink)

How can digitised assets of Galleries, Libraries, Archives and Museums be reused to unlock new value? What are the implications of viewing large-scale cultural heritage data as an economic resource, to build new products and services upon? Drawing upon valuation studies, we reflect on both the theory and practicalities of using mass-digitised heritage content as an economic driver, stressing the need to consider the complexity of commercial-based outcomes within the context of cultural and creative industries. However, we also problematise the (...) act of considering such heritage content as a resource to be exploited for economic growth, in order to inform how we consider, develop, deliver and value mass-digitisation. Our research will be of interest to those wishing to understand a rapidly changing research and innovation landscape, those considering how to engage memory institutions in data-driven activities and those critically evaluating years of mass-digitisation across the heritage sector. (shrink)

This paper explores ethical issues raised by whole slide image-based computational pathology. After briefly giving examples drawn from some recent literature of advances in this field, we consider some ethical problems it might be thought to pose. These arise from the tension between artificial intelligence research—with its hunger for more and more data—and the default preference in data ethics and data protection law for the minimisation of personal data collection and processing; the fact that computational pathology lends itself to kinds (...) of data fusion that go against data ethics norms and some norms of biobanking; the fact that AI methods are esoteric and produce results that are sometimes unexplainable and the fact that computational pathology is particularly dependent on scanning technology manufacturers with interests of their own in profit-making from data collection. We shall suggest that most of these issues are resolvable. (shrink)

A major challenge with the increasing use of Artificial Intelligence applications is to manage the long-term societal impacts of this technology. Two central concerns that have emerged in this respect are that the optimized goals behind the data processing of AI applications usually remain opaque and the energy footprint of their data processing is growing quickly. This study thus explores how much people value the transparency and environmental sustainability of AI using the example of personal AI assistants. The results from (...) a choice-based conjoint analysis with a sample of more than 1.000 respondents from Germany indicate that people hardly care about the energy efficiency of AI; and while they do value transparency through explainable AI, this added value of an application is offset by minor costs. The findings shed light on what kinds of AI people are likely to demand and have important implications for policy and regulation. (shrink)

An Artificial Intelligence algorithm trained on data that reflect racial biases may yield racially biased outputs, even if the algorithm on its own is unbiased. For example, algorithms used to schedule medical appointments in the USA predict that Black patients are at a higher risk of no-show than non-Black patients, though technically accurate given existing data that prediction results in Black patients being overwhelmingly scheduled in appointment slots that cause longer wait times than non-Black patients. This perpetuates racial inequity, in (...) this case lesser access to medical care. This gives rise to one type of Accuracy-Fairness trade-off: preserve the efficiency offered by using AI to schedule appointments or discard that efficiency in order to avoid perpetuating ethno-racial disparities. Similar trade-offs arise in a range of AI applications including others in medicine, as well as in education, judicial systems, and public security, among others. This article presents a framework for addressing such trade-offs where Machine Learning and Optimization components of the algorithm are decoupled. Applied to medical appointment scheduling, our framework articulates four approaches intervening in different ways on different components of the algorithm. Each yields specific results, in one case preserving accuracy comparable to the current state-of-the-art while eliminating the disparity. (shrink)

It is widely acknowledged that high-level AI principles are difficult to translate into practices via explicit rules and design guidelines. Consequently, many AI research and development groups that claim to adopt ethics principles have been accused of unwarranted “ethics washing”. Accordingly, there remains a question as to if and how high-level principles should be expected to influence the development of safe and beneficial AI. In this short commentary I discuss two roles high-level principles might play in AI ethics and governance. (...) The first and most often discussed “start-point” function quickly succumbs to the complaints outlined above. I suggest, however, that a second “cultural influence” function is where the primary value of high-level principles lies. (shrink)

Research into the ethics of artificial intelligence is often categorized into two subareas—robot ethics and machine ethics. Many of the definitions and classifications of the subject matter of these subfields, as found in the literature, are conflated, which I seek to rectify. In this essay, I infer that using the term ‘machine ethics’ is too broad and glosses over issues that the term computational ethics best describes. I show that the subject of inquiry of computational ethics is of great value (...) and indeed is an important frontier in developing ethical artificial intelligence systems. I also show that computational is a distinct, often neglected field in the ethics of AI. In contrast to much of the literature, I argue that the appellation ‘machine ethics’ does not sufficiently capture the entire project of embedding ethics into AI/S, and hence the need for computational ethics. This essay is unique for two reasons; first, it offers a philosophical analysis of the subject of computational ethics that is not found in the literature. Second, it offers a finely grained analysis that shows the thematic distinction among robot ethics, machine ethics and computational ethics. (shrink)

John Danaher and Sven Nyholm have argued that automation, especially of the sort powered by artificial intelligence, poses a threat to meaningful work by diminishing the chances for meaning-conferring workplace achievement, what they call “achievement gaps”. In this paper, I argue that Danaher and Nyholm’s achievement gap thesis suffers from an ambiguity. The weak version of the thesis holds that automation may result in the appearance of achievement gaps, whereas the strong version holds that automation may result on balance loss (...) in possibilities for workplace achievements, i.e., in the appearance of an overall gappier work landscape. Against the strong version of the achievement gap thesis, I characterize situations where automation may result in boons to meaning-conferring workplace achievements: the appearance of what I call achievement spread and achievement swaps. Nevertheless, Danaher and Nyholm are right to worry about some uses of automation whereby human workers become subservient to AI. But these situations are better framed, I argue, as autonomy gaps rather than achievement gaps. (shrink)

This paper provides an in-depth analysis of the impact of AI technologies on the climate crisis beyond their mere resource consumption. To critically examine this impact, I introduce the concept of AI futurism. With this term I capture the ideology behind AI, and argue that this ideology is inherently connected to the climate crisis. This is because AI futurism construes a socio-material environment overly fixated on AI and technological progress, to the extent that it loses sight of the existential threats (...) ahead. In that way, the perceived significance of the planetary reality is softened, and the unsustainable paths charted by the AI industry remain opaque. To make this argument the paper unfolds in two main sections. (1) It delves into the ideology of AI futurism and its discursive impact on the societal perception of the climate crisis. (2) It employs a materialist perspective to elucidate the tangible effects of the AI industry on the climate crisis, and to show what is going on behind the façade of AI futurism. Adding to a critical theory of AI and the climate crisis, this paper starts a realistic conversation about the challenges posed by the intersection of these transformations. (shrink)

Healthcare provision, like many other sectors of society, is undergoing major changes due to the increased use of data-driven methods and technologies. This increased reliance on big data in medicine can lead to shifts in the norms that guide healthcare providers and patients. Continuous critical normative reflection is called for to track such potential changes. This article presents the results of an interview-based study with 20 German and Swiss experts from the fields of medicine, life science research, informatics and humanities (...) of digitalisation. The aim of the study was to explore expert opinions regarding current challenges and opportunities related to data-driven medicine and medical research and to provide a methodological framework for empirically grounded, continuous normative reflection. To this end, we developed a heuristic tool to map and structure empirical findings for normative analysis. Using this tool, our interview material points to a polarisation between individualistic and collectivistic orientated argumentations among experts. The study shows that a multilevel analysis is required to deal with complex normative implications of data-driven approaches in medical research and healthcare. (shrink)

This paper argues that as we move to redefine global bioethics, there is a need to be attentive to the ethical issues associated with the environmental sustainability of data and digital infrastructures in global health systems. We show that these infrastructures have thus far featured little in environmental impact discussions in the context of health, and we use a case study approach of biobanking to illustrate this. We argue that this missing discussion is problematic because biobanks have environmental impacts associated (...) with data and digital infrastructures. We consider several ethical questions to consider these impacts: what ethical work does the concept of environmental sustainability add to the debate; how should this concept be prioritised in decision-making; and who should be responsible for doing so? We call on global bioethics to play a role in advancing this dialogue and addressing these questions. (shrink)

Within the ethical debate on Machine Learning-driven decision support systems (ML_CDSS), notions such as “human in the loop” or “meaningful human control” are often cited as being necessary for ethical legitimacy. In addition, ethical principles usually serve as the major point of reference in ethical guidance documents, stating that conflicts between principles need to be weighed and balanced against each other. Starting from a neo-Kantian viewpoint inspired by Onora O'Neill, this article makes a concrete suggestion of how to interpret the (...) role of the “human in the loop” and to overcome the perspective of rivaling ethical principles in the evaluation of AI in health care. We argue that patients should be perceived as “fellow workers” and epistemic partners in the interpretation of ML_CDSS outputs. We further highlight that a meaningful process of integrating (rather than weighing and balancing) ethical principles is most appropriate in the evaluation of medical AI. (shrink)

How social media impacts the autonomy of its users is a topic of increasing focus. However, much of the literature that explores these impacts fails to engage in depth with the philosophical literature on autonomy. This has resulted in a failure to consider the full range of impacts that social media might have on autonomy. A deeper consideration of these impacts is thus needed, given the importance of both autonomy as a moral concept and social media as a feature of (...) contemporary life. By drawing on this philosophical literature, we argue that autonomy is broadly a matter of developing autonomy competencies, having authentic ends and control over key aspects of your own life, and not being manipulated, coerced, and controlled by others. We show how the autonomy of users of social media can be disrespected and harmed through the control that social media can have over its users’ data, attention, and behaviour. We conclude by discussing various recommendations to better regulate social media. (shrink)

Purpose The purpose of this paper is clearly illustrate this convergence and the prescriptive recommendations that such documents entail. There is a significant amount of research into the ethical consequences of artificial intelligence. This is reflected by many outputs across academia, policy and the media. Many of these outputs aim to provide guidance to particular stakeholder groups. It has recently been shown that there is a large degree of convergence in terms of the principles upon which these guidance documents are (...) based. Despite this convergence, it is not always clear how these principles are to be translated into practice. Design/methodology/approach In this paper, the authors move beyond the high-level ethical principles that are common across the AI ethics guidance literature and provide a description of the normative content that is covered by these principles. The outcome is a comprehensive compilation of normative requirements arising from existing guidance documents. This is not only required for a deeper theoretical understanding of AI ethics discussions but also for the creation of practical and implementable guidance for developers and users of AI. Findings In this paper, the authors therefore provide a detailed explanation of the normative implications of existing AI ethics guidelines but directed towards developers and organisational users of AI. The authors believe that the paper provides the most comprehensive account of ethical requirements in AI currently available, which is of interest not only to the research and policy communities engaged in the topic but also to the user communities that require guidance when developing or deploying AI systems. Originality/value The authors believe that they have managed to compile the most comprehensive document collecting existing guidance which can guide practical action but will hopefully also support the consolidation of the guidelines landscape. The authors’ findings should also be of academic interest and inspire philosophical research on the consistency and justification of the various normative statements that can be found in the literature. (shrink)

Recent work has demonstrated how data-driven AI methods can leverage consumer protection by supporting the automated analysis of legal documents. However, a shortcoming of data-driven approaches is poor explainability. We posit that in this domain useful explanations of classifier outcomes can be provided by resorting to legal rationales. We thus consider several configurations of memory-augmented neural networks where rationales are given a special role in the modeling of context knowledge. Our results show that rationales not only contribute to improve the (...) classification accuracy, but are also able to offer meaningful, natural language explanations of otherwise opaque classifier outcomes. (shrink)

The increasing application of artificial intelligence (AI) to healthcare raises both hope and ethical concerns. Some advanced machine learning methods provide accurate clinical predictions at the expense of a significant lack of explainability. Alex John London has defended that accuracy is a more important value than explainability in AI medicine. In this article, we locate the trade-off between accurate performance and explainable algorithms in the context of distributive justice. We acknowledge that accuracy is cardinal from outcome-oriented justice because it helps (...) to maximize patients’ benefits and optimizes limited resources. However, we claim that the opaqueness of the algorithmic black box and its absence of explainability threatens core commitments of procedural fairness such as accountability, avoidance of bias, and transparency. To illustrate this, we discuss liver transplantation as a case of critical medical resources in which the lack of explainability in AI-based allocation algorithms is procedurally unfair. Finally, we provide a number of ethical recommendations for when considering the use of unexplainable algorithms in the distribution of health-related resources. (shrink)

The increasing demand for transparency in AI has recently come under scrutiny. The question is often posted in terms of “epistemic double standards”, and whether the standards for transparency in AI ought to be higher than, or equivalent to, our standards for ordinary human reasoners. I agree that the push for increased transparency in AI deserves closer examination, and that comparing these standards to our standards of transparency for other opaque systems is an appropriate starting point. I suggest that a (...) more fruitful exploration of this question will involve a different comparison class. We routinely treat judgments made by highly trained experts in specialized fields as fair or well-grounded even though—by the nature of expert/layperson division of epistemic labor—an expert will not be able to provide an explanation of the reasoning behind these judgments that makes sense to most other people. Regardless, laypeople are thought to be acting reasonably—and ethically—in deferring to the judgments of experts that concern their areas of specialization. I suggest that we reframe our question regarding the appropriate standards of transparency in AI as one that asks when, why, and to what degree it would be ethical to accept opacity in AI. I argue that our epistemic relation to certain opaque AI technology may be relevantly similar to the layperson’s epistemic relation to the expert in certain respects, such that the successful expert/layperson division of epistemic labor can serve as a blueprint for the ethical use of opaque AI. (shrink)

El aumento considerable de la capacidad de la inteligencia artificial (IA) implica un alto consumo de recursos energéticos. La situación ambiental actual, caracterizada por la acuciante degradación de ecosistemas y la ruptura del equilibrio, exige tomar medidas en diversos ámbitos. La IA no puede quedar al margen, y aunque es empleada para objetivos de sostenibilidad, debe plantearse como sostenible en términos integrales. La propuesta de una inteligencia artificial sostenible se argumenta a partir de una evaluación ética constructiva, donde la inclusión (...) y la participación de los grupos de interés representan dos elementos fundamentales. (shrink)

Instituciones políticas como la Comisión Europea y el Gobierno de España han manifestadosu interés y predisposición para sentar las bases de una gobernanza ética de la Inteligencia Artificial (IA). En particular, han planteado el impulso de una Inteligencia Artificial confiable a través de unconjunto de directrices y estrategias. A pesar del beneficio que reportan estas iniciativas políticas, no es posible apreciar en su conjunto una estrategia educativa específica que contribuya a la generación de un ecosistema ético de IA fundamentado en (...) la confianza. En ese sentido, el objetivo principal de este trabajo consiste en explicar que la enseñanza de la ética aplicada en los estudios de ingeniería constituye una apuesta para el fortalecimiento profesional, la gobernanza ética yuna investigación e innovación responsables. Para alcanzar este objetivo, en primer lugar, serándetallados algunos de los aspectos esenciales que dotan de un valor significativo a la enseñanza de la ética aplicada, subrayando la responsabilidad social que presentan las universidades en este terreno. En segundo lugar, se mostrará una clasificación de las asignaturas de ética relacionadas con la ingeniería en el entorno de las universidades públicas españolas. Y, en tercer lugar, se ofrecerá un marco teórico enraizado en la ética discursiva, promoviendo una mirada cívica en el contexto educativo de las profesiones. (shrink)

There is widespread agreement that there should be a principle requiring that artificial intelligence be ‘explicable’. Microsoft, Google, the World Economic Forum, the draft AI ethics guidelines for the EU commission, etc. all include a principle for AI that falls under the umbrella of ‘explicability’. Roughly, the principle states that “for AI to promote and not constrain human autonomy, our ‘decision about who should decide’ must be informed by knowledge of how AI would act instead of us” :689–707, 2018). There (...) is a strong intuition that if an algorithm decides, for example, whether to give someone a loan, then that algorithm should be explicable. I argue here, however, that such a principle is misdirected. The property of requiring explicability should attach to a particular action or decision rather than the entity making that decision. It is the context and the potential harm resulting from decisions that drive the moral need for explicability—not the process by which decisions are reached. Related to this is the fact that AI is used for many low-risk purposes for which it would be unnecessary to require that it be explicable. A principle requiring explicability would prevent us from reaping the benefits of AI used in these situations. Finally, the explanations given by explicable AI are only fruitful if we already know which considerations are acceptable for the decision at hand. If we already have these considerations, then there is no need to use contemporary AI algorithms because standard automation would be available. In other words, a principle of explicability for AI makes the use of AI redundant. (shrink)

Over the past few years, there has been a proliferation of artificial intelligence strategies, released by governments around the world, that seek to maximise the benefits of AI and minimise potential harms. This article provides a comparative analysis of the European Union and the United States’ AI strategies and considers the visions of a ‘Good AI Society’ that are forwarded in key policy documents and their opportunity costs, the extent to which the implementation of each vision is living up to (...) stated aims and the consequences that these differing visions of a ‘Good AI Society’ have for transatlantic cooperation. The article concludes by comparing the ethical desirability of each vision and identifies areas where the EU, and especially the US, need to improve in order to achieve ethical outcomes and deepen cooperation. (shrink)

Artificial Intelligence (AI) seems to be impacting all industry sectors, while becoming a motor for innovation. The diffusion of AI from the civilian sector to the defense sector, and AI’s dual-use potential has drawn attention from security and ethics scholars. With the publication of the ethical guideline Trustworthy AI by the European Union (EU), normative questions on the application of AI have been further evaluated. In order to draw conclusions on Trustworthy AI as a point of reference for responsible research (...) and development (R&D), we approach the diffusion of AI across both civilian and military spheres in the EU. We capture the extent of technological diffusion and derive European and German patent citation networks. Both networks indicate a low degree of diffusion of AI between civilian and defense sectors. A qualitative investigation of project descriptions of a research institute’s work in both civilian and military fields shows that military AI applications stress accuracy or robustness, while civilian AI reflects a focus on human-centric values. Our work represents a first approach by linking processes of technology diffusion with normative evaluations of R&D. (shrink)

Despite the bombardment of AI ethics frameworks (AIEFs) published in the last decade, it is unclear which of the many have been adopted in the industry. What is more, the sheer volume of AIEFs without a clear demonstration of their effectiveness makes it difficult for businesses to select which framework they should adopt. As a first step toward addressing this problem, we employed four different existing frameworks to assess AI ethics concerns of a real-world AI system. We compared the experience (...) of applying the AIEFs from the perspective of (a) a third-party auditor conducting an AI ethics risk assessment for the company, and (b) the company receiving the audit outcomes. Our results suggest that the feel-good factor of doing an assessment is common across the AIEFs that can take anywhere between 1.5 and 20 h to complete. However, each framework provides different benefits (e.g., issue discovery vs. issue monitoring) and is likely best used in conjunction with one another at different stages of an AI development process. As such, we call on the AI ethics community to better specify the suitability and expected benefits of existing frameworks to enable better adoption of AI ethics practice in the industry. (shrink)

Artificial intelligence and automation have become pervasive in news media, influencing journalism from news gathering to news distribution. As algorithms are increasingly determining editorial decisions, specific concerns have been raised with regard to the responsible and accountable use of AI-driven tools by news media, encompassing new regulatory and ethical questions. This contribution aims to analyze whether and to what extent the use of AI technology in news media and journalism is currently regulated and debated within the European Union and the (...) Council of Europe. Through a document analysis of official policy documents, combined with a data mining approach and an inductive thematic analysis, the study looks at how news media are dealt with, in particular regarding their responsibilities towards their users and society. The findings show that regulatory frameworks about AI rarely include media, but if they do, they associate them with issues such as disinformation, data, and AI literacy, as well as diversity, plurality, and social responsibility. (shrink)

Assistive systems based on Artificial Intelligence (AI) are bound to reshape decision-making in all areas of society. One of the most intricate challenges arising from their implementation in high-stakes environments such as medicine concerns their frequently unsatisfying levels of explainability, especially in the guise of the so-called black-box problem: highly successful models based on deep learning seem to be inherently opaque, resisting comprehensive explanations. This may explain why some scholars claim that research should focus on rendering AI systems understandable, rather (...) than explainable. Yet, there is a grave lack of agreement concerning these terms in much of the literature on AI. We argue that the seminal distinction made by the philosopher and physician Karl Jaspers between different types of explaining and understanding in psychopathology can be used to promote greater conceptual clarity in the context of Machine Learning (ML). Following Jaspers, we claim that explaining and understanding constitute multi-faceted epistemic approaches that should not be seen as mutually exclusive, but rather as complementary ones as in and of themselves they are necessarily limited. Drawing on the famous example of Watson for Oncology we highlight how Jaspers’ methodology translates to the case of medical AI. Classical considerations from the philosophy of psychiatry can therefore inform a debate at the centre of current AI ethics, which in turn may be crucial for a successful implementation of ethically and legally sound AI in medicine. (shrink)

The justification and rationality of this paper is to present some fundamental principles, theories, and concepts that we believe moulds the nucleus of a good artificial intelligence (AI) society. The morally accepted significance and utilitarian concerns that stems from the inception and realisation of an AI’s structural foundation are displayed in this study. This paper scrutinises the structural foundation, fundamentals, and cardinal righteous remonstrations, as well as the gaps in mechanisms towards novel prospects and perils in determining resilient fundamentals, accountability, (...) and AI’s convoluted and responsible implications. We outline a number of salient and practical benefits, in which to place moral norms within the mise en scène of AI, to delineate the rudimentary ethical dilemmas and decorous directions within the realms of AI. (shrink)