Artificial intelligence (AI) has become nearly ubiquitous in modern society, from components of mobile applications to medical support systems, and everything in between. In societally impactful systems imbued with AI, there has been increasing concern related to opaque AI, that is, artificial intelligence where it is unclear how or why certain decisions are reached. This has led to a recent boom in research on “explainable AI” (XAI), or approaches to making AI more explainable and understandable to human users. In the (...) military domain, numerous bodies have argued that autonomous and AI-enabled weapon systems ought not incorporate unexplainable AI, with the International Committee of the Red Cross and the United States Department of Defense both explicitly including explainability as a relevant factor in the development and use of such systems. In this article, I present a cautiously critical assessment of this view, arguing that explainability will be irrelevant for many current and near-future autonomous systems in the military (which do not incorporate any AI), that it will be trivially incorporated into most military systems which do possess AI (as these generally possess simpler AI systems), and that for those systems with genuinely opaque AI, explainability will prove to be of more limited value than one might imagine. In particular, I argue that explainability, while indeed a virtue in design, is a virtue aimed primarily at designers and troubleshooters of AI-enabled systems, but is far less relevant for users and handlers actually deploying these systems. I further argue that human–machine teaming is a far more important element of responsibly using AI for military purposes, adding that explainability may undermine efforts to improve human–machine teamings by creating aprima faciesense that the AI, due to its explainability, may be utilized with little (or less) potential for mistakes. I conclude by clarifying that the arguments are not against XAI in the military, but are instead intended as a caution against over-inflating the value of XAI in this domain, or ignoring the limitations and potential pitfalls of this approach. (shrink)

We argue that the trend toward providing users with feasible and actionable explanations of AI decisions—known as recourse explanations—comes with ethical downsides. Specifically, we argue that recourse explanations face several conceptual pitfalls and can lead to problematic explanation hacking, which undermines their ethical status. As an alternative, we advocate that explanations of AI decisions should aim at understanding.

As systems based on opaque Artificial Intelligence (AI) continue to flourish in diverse real-world applications, understanding these black box models has become paramount. In response, Explainable AI (XAI) has emerged as a field of research with practical and ethical benefits across various domains. This paper not only highlights the advancements in XAI and its application in real-world scenarios but also addresses the ongoing challenges within XAI, emphasizing the need for broader perspectives and collaborative efforts. We bring together experts from diverse (...) fields to identify open problems, striving to synchronize research agendas and accelerate XAI in practical applications. By fostering collaborative discussion and interdisciplinary cooperation, we aim to propel XAI forward, contributing to its continued success. Our goal is to put forward a comprehensive proposal for advancing XAI. To achieve this goal, we present a manifesto of 27 open problems categorized into nine categories. These challenges encapsulate the complexities and nuances of XAI and offer a road map for future research. For each problem, we provide promising research directions in the hope of harnessing the collective intelligence of interested stakeholders. (shrink)

The ongoing explosion of interest in artificial intelligence is fueled in part by recently developed techniques in machine learning. Those techniques allow automated systems to process huge amounts of data, utilizing mathematical methods that depart from traditional statistical approaches, and resulting in impressive advancements in our ability to make predictions and uncover correlations across a host of interesting domains. But as is now widely discussed, the way that those systems arrive at their outputs is often opaque, even to the experts (...) who design and deploy them. Is it morally problematic to make use of opaque automated methods when making high-stakes decisions, like whether to issue a loan to an applicant, or whether to approve a parole request? Many scholars answer in the affirmative. However, there is no widely accepted explanation for why transparent systems are morally preferable to opaque systems. We argue that the use of automated decision-making systems sometimes violates duties of consideration that are owed by decision-makers to decision-subjects, duties that are both epistemic and practical in character. Violations of that kind generate a weighty consideration against the use of opaque decision systems. In the course of defending our approach, we show that it is able to address three major challenges sometimes leveled against attempts to defend the moral import of transparency in automated decision-making. (shrink)

Explainable artificial intelligence (XAI) seeks to produce explanations for those machine learning methods which are deemed opaque. However, there is considerable disagreement about what this means and how to achieve it. Authors disagree on what should be explained (topic), to whom something should be explained (stakeholder), how something should be explained (instrument), and why something should be explained (goal). In this paper, I employ insights from means-end epistemology to structure the field. According to means-end epistemology, different means ought to be (...) rationally adopted to achieve different epistemic ends. Applied to XAI, different topics, stakeholders, and goals thus require different instruments. I call this the means-end account of XAI. The means-end account has a descriptive and a normative component: on the one hand, I show how the specific means-end relations give rise to a taxonomy of existing contributions to the field of XAI; on the other hand, I argue that the suitability of XAI methods can be assessed by analyzing whether they are prescribed by a given topic, stakeholder, and goal. (shrink)

This paper examines racial discrimination and algorithmic bias in predictive policing algorithms (PPAs), an emerging technology designed to predict threats and suggest solutions in law enforcement. We first describe what discrimination is in a case study of Chicago’s PPA. We then explain their causes with Broadbent’s contrastive model of causation and causal diagrams. Based on the cognitive science literature, we also explain why fairness is not an objective truth discoverable in laboratories but has context-sensitive social meanings that need to be (...) negotiated through democratic processes. With the above analysis, we next predict why some recommendations given in the bias reduction literature are not as effective as expected. Unlike the cliché highlighting equal participation for all stakeholders in predictive policing, we emphasize power structures to avoid hermeneutical lacunae. Finally, we aim to control PPA discrimination by proposing a governance solution—a framework of a social safety net. (shrink)

The need for fair and just AI is often related to the possibility of understanding AI itself, in other words, of turning an opaque box into a glass box, as inspectable as possible. Transparency and explainability, however, pertain to the technical domain and to philosophy of science, thus leaving the ethics and epistemology of AI largely disconnected. To remedy this, we propose an integrated approach premised on the idea that a glass-box epistemology should explicitly consider how to incorporate values and (...) other normative considerations, such as intersectoral vulnerabilities, at critical stages of the whole process from design and implementation to use and assessment. To connect ethics and epistemology of AI, we perform a double shift of focus. First, we move from trusting the output of an AI system to trusting the process that leads to the outcome. Second, we move from expert assessment to more inclusive assessment strategies, aiming to facilitate expert and non-expert assessment. Together, these two moves yield a framework usable for experts and non-experts when they inquire into relevant epistemological and ethical aspects of AI systems. We dub our framework ‘epistemology-cum-ethics’ to signal the equal importance of both aspects. We develop it from the vantage point of the designers: how to create the conditions to internalize values into the whole process of design, implementation, use, and assessment of an AI system, in which values (epistemic and non-epistemic) are explicitly considered at each stage and inspectable by every salient actor involved at any moment. (shrink)

Many AI systems that make important decisions are black boxes: how they function is opaque even to their developers. This is due to their high complexity and to the fact that they are trained rather than programmed. Efforts to alleviate the opacity of black box systems are typically discussed in terms of transparency, interpretability, and explainability. However, there is little agreement about what these key concepts mean, which makes it difficult to adjudicate the success or promise of opacity alleviation methods. (...) I argue for a unified account of these key concepts that treats the concept of understanding as fundamental. This allows resources from the philosophy of science and the epistemology of understanding to help guide opacity alleviation efforts. A first significant benefit of this understanding account is that it defuses one of the primary, in-principle objections to post hoc explainable AI (XAI) methods. This “rationalization objection” argues that XAI methods provide mere rationalizations rather than genuine explanations. This is because XAI methods involve using a separate “explanation” system to approximate the original black box system. These explanation systems function in a completely different way than the original system, yet XAI methods make inferences about the original system based on the behavior of the explanation system. I argue that, if we conceive of XAI methods as idealized scientific models, this rationalization worry is dissolved. Idealized scientific models misrepresent their target phenomena, yet are capable of providing significant and genuine understanding of their targets. (shrink)

Advancements in machine learning have fuelled the popularity of using AI decision algorithms in procedures such as bail hearings, medical diagnoses and recruitment. Academic articles, policy texts, and popularizing books alike warn that such algorithms tend to be opaque: they do not provide explanations for their outcomes. Building on a causal account of transparency and opacity as well as recent work on the value of causal explanation, I formulate a moral concern for opaque algorithms that is yet to receive a (...) systematic treatment in the literature: when such algorithms are used in life-changing decisions, they can obstruct us from effectively shaping our lives according to our goals and preferences, thus undermining our autonomy. I argue that this concern deserves closer attention as it furnishes the call for transparency in algorithmic decision-making with both new tools and new challenges. (shrink)

Philosophy Compass, Volume 17, Issue 6, June 2022.

We argue that explainable artificial intelligence (XAI), specifically reason-giving XAI, often constitutes the most suitable way of ensuring that someone can properly be held responsible for decisions that are based on the outputs of artificial intelligent (AI) systems. We first show that, to close moral responsibility gaps (Matthias 2004), often a human in the loop is needed who is directly responsible for particular AI-supported decisions. Second, we appeal to the epistemic condition on moral responsibility to argue that, in order to (...) be responsible for her decision, the human in the loop has to have an explanation available of the system’s recommendation. Reason explanations are especially well-suited to this end and we examine whether – and how – it might be possible to make such explanations fit with AI systems. We support our claims by focusing on a case of disagreement between human in the loop and AI system. (shrink)

Explanations are conceived to ensure the trustworthiness of AI systems. Yet, relying solemnly on algorithmic solutions, as provided by explainable artificial intelligence (XAI), might fall short to account for sociotechnical risks jeopardizing their factuality and informativeness. To mitigate these risks, we delve into the complex landscape of ethical risks surrounding XAI systems and their generated explanations. By employing a literature review combined with rigorous thematic analysis, we uncover a diverse array of technical risks tied to the robustness, fairness, and evaluation (...) of XAI systems. Furthermore, we address a broader range of contextual risks jeopardizing their security, accountability, reception alongside other cognitive, social, and ethical concerns of explanations. We advance a multi-layered risk assessment framework, where each layer advances strategies for practical intervention, management, and documentation of XAI systems within organizations. Recognizing the theoretical nature of the framework advanced, we discuss it in a conceptual case study. For the XAI community, our multifaceted investigation represents a path to practically address XAI risks while enriching our understanding of the ethical ramifications of incorporating XAI in decision-making processes. (shrink)

When artificial intelligence (AI) is used to make high-stakes decisions, some worry that this will create a morally troubling responsibility gap—that is, a situation in which nobody is morally responsible for the actions and outcomes that result. Since the responsibility gap might be thought to result from individuals lacking knowledge of the future behavior of AI systems, it can be and has been suggested that deploying explainable artificial intelligence (XAI) techniques will help us to avoid it. These techniques provide humans (...) with certain forms of understanding of the systems in question. In this paper, I consider whether existing XAI techniques can indeed close the responsibility gap. I identify a number of significant limits to their ability to do so. Ensuring that responsibility for AI-assisted outcomes is maintained may require using different techniques in different circumstances, and potentially also developing new techniques that can avoid each of the issues identified. (shrink)

Artificial intelligent (AI) systems that perform image classification tasks are being used to great success in many application contexts. However, many of these systems are opaque, even to experts. This lack of understanding can be problematic for ethical, legal, or practical reasons. The research field Explainable AI (XAI) has therefore developed several approaches to explain image classifiers. The hope is to bring about understanding, e.g., regarding why certain images are classified as belonging to a particular target class. Most of these (...) approaches use visual explanations. Drawing on Elgin’s work (True enough. MIT Press, Cambridge, 2017), I argue that analyzing what those explanations exemplify can help to assess their suitability for producing understanding. More specifically, I suggest to distinguish between two forms of examples according to their suitability for producing understanding. I call these forms samples and exemplars, respectively. Samples are prone to misinterpretation and thus carry the risk of leading to misunderstanding. Exemplars, by contrast, are intentionally designed or chosen to meet contextual requirements and to mitigate the risk of misinterpretation. They are thus preferable for bringing about understanding. By reviewing several XAI approaches directed at image classifiers, I show that most of them explain with samples. If my analysis is correct, it will be beneficial if such explainability methods use explanations that qualify as exemplars. (shrink)

With its potential to contribute to the ethical governance of AI, eXplainable AI (XAI) research frequently asserts its relevance to ethical considerations. Yet, the substantiation of these claims with rigorous ethical analysis and reflection remains largely unexamined. This contribution endeavors to scrutinize the relationship between XAI and ethical considerations. By systematically reviewing research papers mentioning ethical terms in XAI frameworks and tools, we investigate the extent and depth of ethical discussions in scholarly research. We observe a limited and often superficial (...) engagement with ethical theories, with a tendency to acknowledge the importance of ethics, yet treating it as a monolithic and not contextualized concept. Our findings suggest a pressing need for a more nuanced and comprehensive integration of ethics in XAI research and practice. To support this, we propose to critically reconsider transparency and explainability in regards to ethical considerations during XAI systems design while accounting for ethical complexity in practice. As future research directions, we point to the promotion of interdisciplinary collaborations and education, also for underrepresented ethical perspectives. Such ethical grounding can guide the design of ethically robust XAI systems, aligning technical advancements with ethical considerations. (shrink)

This paper analyses whether current explainable AI (XAI) techniques can help to address taxpayer concerns about the use of AI in taxation. As tax authorities around the world increase their use of AI-based techniques, taxpayers are increasingly at a loss about whether and how the ensuing decisions follow the procedures required by law and respect their substantive rights. The use of XAI has been proposed as a response to this issue, but it is still an open question whether current XAI (...) techniques are enough to meet existing legal requirements. The paper approaches this question in the context of a case study: a prototype tax fraud detector trained on an anonymized dataset of real-world cases handled by the Buenos Aires (Argentina) tax authority. The decisions produced by this detector are explained through the use of various classification methods, and the outputs of these explanation models are evaluated on their explanatory power and on their compliance with the legal obligation that tax authorities provide the rationale behind their decision-making. We conclude the paper by suggesting technical and legal approaches for designing explanation mechanisms that meet the needs of legal explanation in the tax domain. (shrink)

This paper discusses some societal implications of the most recent and publicly discussed application of advanced machine learning techniques: generative AI models, such as ChatGPT (text generation) and DALL-E (text-to-image generation). The aim is to shift attention away from conceptual disputes, e.g. regarding their level of intelligence and similarities/differences with human performance, to focus instead on practical problems, pertaining the impact that these technologies might have (and already have) on human societies. After a preliminary clarification of how generative AI works (...) (Sect. 1), the paper discusses what kind of transparency ought to be required for such technologies and for the business model behind their commercial exploitation (Sect. 2), what is the role of user-generated data in determining their performance and how it should inform the redistribution of the resulting benefits (Sect. 3), the best way of integrating generative AI systems in the creative job market and how to properly negotiate their role in it (Sect. 4), and what kind of “cognitive extension” offered by these technologies we ought to embrace, and what type we should instead resist and monitor (Sect. 5). The last part of the paper summarizes the main conclusions of this analysis, also marking its distance from other, more apocalyptic approaches to the dangers of AI for human society. (shrink)

Recent work in explainable artificial intelligence (XAI) attempts to render opaque AI systems understandable through a divide-and-conquer strategy. However, this fails to illuminate how trained AI systems work as a whole. Precisely this kind of functional understanding is needed, though, to satisfy important societal desiderata such as safety. To remedy this situation, we argue, AI researchers should seek mechanistic interpretability, viz. apply coordinated discovery strategies familiar from the life sciences to uncover the functional organisation of complex AI systems. Additionally, theorists (...) should accommodate for the unique costs and benefits of such strategies in their portrayals of XAI research. (shrink)

Legislation and ethical guidelines around the globe call for effective human oversight of AI-based systems in high-risk contexts – that is oversight that reliably reduces the risks otherwise associated with the use of AI-based systems. Such risks may relate to the imperfect accuracy of systems (e.g., inaccurate classifications) or to ethical concerns (e.g., unfairness of outputs). Given the significant role that human oversight is expected to play in the operation of AI-based systems, it is crucial to better understand the conditions (...) for effective human oversight. We argue that the reliable detection of errors (as an umbrella term for inaccuracies and unfairness) is crucial for effective human oversight. We then propose that Signal Detection Theory (SDT) offers a promising framework for better understanding what affects people’s sensitivity (i.e., how well they are able to detect errors) and response bias (i.e., the tendency to report errors given a perceived evidence of an error) in detecting errors. Whereas an SDT perspective on the detection of inaccuracies is straightforward, we demonstrate its broader applicability by detailing the specifics for an SDT perspective on unfairness detection, including the need to choose a standard for (un)fairness. Additionally, we illustrate that an SDT perspective helps to better understand the conditions for effective error detection by showing examples of task-, system-, and person-related factors that may affect the sensitivity and response bias of humans tasked with detecting unfairness associated with the use of AI-based systems. Finally, we discuss future research directions for an SDT perspective on error detection. (shrink)