Generative AI programs can produce high-quality written and visual content that may be used for good or ill. We argue that a credit–blame asymmetry arises for assigning responsibility for these outputs and discuss urgent ethical and policy implications focused on large-scale language models.

Artificial intelligence (AI) and its introduction into clinical pathways presents an array of ethical issues that are being discussed in the JME. 1–7 The development of AI technologies that can produce text that will pass plagiarism detectors 8 and are capable of appearing to be written by a human author 9 present new issues for medical ethics. One set of worries concerns authorship and whether it will now be possible to know that an author or student in fact produced submitted (...) work. That seems likely to be a general worry for secondary and higher education, as well as for all academic journals. Thus far generative AI chatbots do not seem to be able to produce a fully referenced and well-argued ethics article, but they probably could generate a blog or student essay that would... (shrink)

Large-scale language technologies are increasingly used in various forms of communication with humans across different contexts. One particular use case for these technologies is conversational agents, which output natural language text in response to prompts and queries. This mode of engagement raises a number of social and ethical questions. For example, what does it mean to align conversational agents with human norms or values? Which norms or values should they be aligned with? And how can this be accomplished? In this (...) paper, we propose a number of steps that help answer these questions. We start by developing a philosophical analysis of the building blocks of linguistic communication between conversational agents and human interlocutors. We then use this analysis to identify and formulate ideal norms of conversation that can govern successful linguistic communication between humans and conversational agents. Furthermore, we explore how these norms can be used to align conversational agents with human values across a range of different discursive domains. We conclude by discussing the practical implications of our proposal for the design of conversational agents that are aligned with these norms and values. (shrink)

We are at a turning point in the debate on the ethics of Artificial Intelligence (AI) because we are witnessing the rise of general-purpose AI text agents such as GPT-3 that can generate large-scale highly refined content that appears to have been written by a human. Yet, a discussion on the ethical issues related to the blurring of the roles between humans and machines in the production of content in the business arena is lacking. In this conceptual paper, drawing on (...) agenda setting theory and stakeholder theory, we challenge the current debate on the ethics of AI and aim to stimulate studies that develop research around three new challenges of AI text agents: automated mass manipulation and disinformation (i.e., fake agenda problem), massive low-quality content production (i.e., lowest denominator problem) and the creation of a growing buffer in the communication between stakeholders (i.e., the mediation problem). (shrink)

The ethics of artificial intelligence, or AI ethics, is a rapidly growing field, and rightly so. While the range of issues and groups of stakeholders concerned by the field of AI ethics is expanding, with speculation about whether it extends even to the machines themselves, there is a group of sentient beings who are also affected by AI, but are rarely mentioned within the field of AI ethics—the nonhuman animals. This paper seeks to explore the kinds of impact AI has (...) on nonhuman animals, the severity of these impacts, and their moral implications. We hope that this paper will facilitate the development of a new field of philosophical and technical research regarding the impacts of AI on animals, namely, the ethics of AI as it affects nonhuman animals. (shrink)

Many ethics initiatives have stipulated sets of principles and standards for good technology development in the AI sector. However, several AI ethics researchers have pointed out a lack of practical realization of these principles. Following that, AI ethics underwent a practical turn, but without deviating from the principled approach. This paper proposes a complementary to the principled approach that is based on virtue ethics. It defines four “basic AI virtues”, namely justice, honesty, responsibility and care, all of which represent specific (...) motivational settings that constitute the very precondition for ethical decision making in the AI field. Moreover, it defines two “second-order AI virtues”, prudence and fortitude, that bolster achieving the basic virtues by helping with overcoming bounded ethicality or hidden psychological forces that can impair ethical decision making and that are hitherto disregarded in AI ethics. Lastly, the paper describes measures for successfully cultivating the mentioned virtues in organizations dealing with AI research and development. (shrink)

The proposed European Artificial Intelligence Act (AIA) is the first attempt to elaborate a general legal framework for AI carried out by any major global economy. As such, the AIA is likely to become a point of reference in the larger discourse on how AI systems can (and should) be regulated. In this article, we describe and discuss the two primary enforcement mechanisms proposed in the AIA: the _conformity assessments_ that providers of high-risk AI systems are expected to conduct, and (...) the _post-market monitoring plans_ that providers must establish to document the performance of high-risk AI systems throughout their lifetimes. We argue that the AIA can be interpreted as a proposal to establish a Europe-wide ecosystem for conducting AI auditing, albeit in other words. Our analysis offers two main contributions. First, by describing the enforcement mechanisms included in the AIA in terminology borrowed from existing literature on AI auditing, we help providers of AI systems understand how they can prove adherence to the requirements set out in the AIA in practice. Second, by examining the AIA from an auditing perspective, we seek to provide transferable lessons from previous research about how to refine further the regulatory approach outlined in the AIA. We conclude by highlighting seven aspects of the AIA where amendments (or simply clarifications) would be helpful. These include, above all, the need to translate vague concepts into verifiable criteria and to strengthen the institutional safeguards concerning conformity assessments based on internal checks. (shrink)

This article reports the findings of AI4People, an Atomium—EISMD initiative designed to lay the foundations for a “Good AI Society”. We introduce the core opportunities and risks of AI for society; present a synthesis of five ethical principles that should undergird its development and adoption; and offer 20 concrete recommendations—to assess, to develop, to incentivise, and to support good AI—which in some cases may be undertaken directly by national or supranational policy makers, while in others may be led by other (...) stakeholders. If adopted, these recommendations would serve as a firm foundation for the establishment of a Good AI Society. (shrink)

Traditional theories of moral development emphasize the role of controlled cognition in mature moral judgment, while a more recent trend emphasizes intuitive and emotional processes. Here we test a dual-process theory synthesizing these perspectives. More specifically, our theory associates utilitarian moral judgment (approving of harmful actions that maximize good consequences) with controlled cognitive processes and associates non-utilitarian moral judgment with automatic emotional responses. Consistent with this theory, we find that a cognitive load manipulation selectively interferes with utilitarian judgment. This interference (...) effect provides direct evidence for the influence of controlled cognitive processes in moral judgment, and utilitarian moral judgment more specifically. (shrink)

Recent research in moral psychology highlights the role of emotion and intuition in moral judgment. In the wake of these findings, the role and significance of moral reasoning remain uncertain. In this article, we distinguish among different kinds of moral reasoning and review evidence suggesting that at least some kinds of moral reasoning play significant roles in moral judgment, including roles in abandoning moral intuitions in the absence of justifying reasons, applying both deontological and utilitarian moral principles, and counteracting automatic (...) tendencies toward bias that would otherwise dominate behavior. We argue that little is known about the psychology of moral reasoning and that it may yet prove to be a potent social force. (shrink)

Obtaining informed consent from patients prior to a medical or surgical procedure is a fundamental part of safe and ethical clinical practice. Currently, it is routine for a significant part of the consent process to be delegated to members of the clinical team not performing the procedure (eg, junior doctors). However, it is common for consent-taking delegates to lack sufficient time and clinical knowledge to adequately promote patient autonomy and informed decision-making. Such problems might be addressed in a number of (...) ways. One possible solution to this clinical dilemma is through the use of conversational artificial intelligence using large language models (LLMs). There is considerable interest in the potential benefits of such models in medicine. For delegated procedural consent, LLM could improve patients’ access to the relevant procedural information and therefore enhance informed decision-making.In this paper, we first outline a hypothetical example of delegation of consent to LLMs prior to surgery. We then discuss existing clinical guidelines for consent delegation and some of the ways in which current practice may fail to meet the ethical purposes of informed consent. We outline and discuss the ethical implications of delegating consent to LLMs in medicine concluding that at least in certain clinical situations, the benefits of LLMs potentially far outweigh those of current practices. (shrink)

How can one build AI systems such that they pursue the goals their designers want them to pursue? This is the alignment problem. Numerous authors have raised concerns that, as research advances and systems become more powerful over time, misalignment might lead to catastrophic outcomes, perhaps even to the extinction or permanent disempowerment of humanity. In this paper, I analyze the severity of this risk based on current instances of misalignment. More specifically, I argue that contemporary large language models and (...) game-playing agents are sometimes misaligned. These cases suggest that misalignment tends to have a variety of features: misalignment can be hard to detect, predict and remedy, it does not depend on a specific architecture or training paradigm, it tends to diminish a system’s usefulness and it is the default outcome of creating AI via machine learning. Subsequently, based on these features, I show that the risk of AI alignment magnifies with respect to more capable systems. Not only might more capable systems cause more harm when misaligned, aligning them should be expected to be more difficult than aligning current AI. (shrink)

In this article I discuss the ethical ramifications for medical ethics training of the availability of large language models (LLMs) for medical students. My focus is on the practical ethical consequences for what we should expect of medical students in terms of medical professionalism and ethical reasoning, and how this can be tested in a context where LLMs are relatively easy available. If we continue to expect ethical competences of medical professionalism of future physicians, how much – if at all (...) – should we worry that such generative AI may compromise adequate testing of medical students’ abilities in this regard? I mainly focus on assessment methods based on written assignments of the ‘student paper’ type and consider whether LLMs make it unfeasible for assessors to gauge whether output is student-generated or ‘machine-generated’ and, if so, whether this is a problem. My take on this research question unfolds in three interwoven arguments, claiming that the advent of LLMs may offer a momentum (i) to reaffirm the importance of context-sensitive interpretation and specification of ethical principles in medical ethics training, (ii) to provide more supportive circumstances to assessors to allow them to meet scoring demands entailed by the importance that is placed on medical professionalism, and (iii) to complement written assignments with verbal (group) discussion to train and test students’ skills to habitually recognize that ‘moral solutions’ can be normatively questioned and specified from various perspectives. (shrink)

We discuss potential biorisks from large language models (LLMs). AI assistants based on LLMs such as ChatGPT have been shown to significantly reduce barriers to entry for actors wishing to synthesize dangerous, potentially novel pathogens and chemical weapons. The harms from deploying such bioagents could be further magnified by AI-assisted misinformation. We endorse several policy responses to these dangers, including prerelease evaluations of biomedical AIs by subject-matter experts, enhanced surveillance and lab screening procedures, restrictions on AI training data, and access (...) controls on biomedical AI. We conclude with a suggestion about future research directions in bioethics. (shrink)

When discussing AI alignment, we usually refer to the problem of teaching or training advanced autonomous AI systems to make decisions that are aligned with human values or preferences. Proponents of this approach believe it can be employed as means to stay in control over sophisticated intelligent systems, thus avoiding certain existential risks. We identify three general obstacles on the path to implementation of value alignment: a technological/technical obstacle, a normative obstacle, and a calibration problem. Presupposing, for the purposes of (...) this discussion, that the technical and normative problems are solved, we focus on the problem of how to calibrate a system, for a specific value, to be on a specific location within a spectrum stretching between righteous and normal or average human behavior. Calibration, or more specifically mis-calibration, also raises the issue of trustworthiness. If we cannot trust AI systems to perform tasks the way we intended, we would not use them on our roads and at our homes. In an era where we strive to construct autonomous machines endowed with common sense, reasoning abilities and a connection to the world, so they would be able to act in alignment with human values, such mis-calibrations can make the difference between trustworthy and untrustworthy systems. (shrink)

Many researchers and intellectuals warn about extreme risks from artificial intelligence. However, these warnings typically came without systematic arguments in support. This paper provides an argument that AI will lead to the permanent disempowerment of humanity, e.g. human extinction, by 2100. It rests on four substantive premises which it motivates and defends: first, the speed of advances in AI capability, as well as the capability level current systems have already reached, suggest that it is practically possible to build AI systems (...) capable of disempowering humanity by 2100. Second, due to incentives and coordination problems, if it is possible to build such AI, it will be built. Third, since it appears to be a hard technical problem to build AI which is aligned with the goals of its designers, and many actors might build powerful AI, misaligned powerful AI will be built. Fourth, because disempowering humanity is useful for a large range of misaligned goals, such AI will try to disempower humanity. If AI is capable of disempowering humanity and tries to disempower humanity by 2100, then humanity will be disempowered by 2100. This conclusion has immense moral and prudential significance. (shrink)

In this article, we discuss ethical issues related to using and disclosing artificial intelligence (AI) tools, such as ChatGPT and other systems based on large language models (LLMs), to write or edit scholarly manuscripts. Some journals, such as Science, have banned the use of LLMs because of the ethical problems they raise concerning responsible authorship. We argue that this is not a reasonable response to the moral conundrums created by the use of LLMs because bans are unenforceable and would encourage (...) undisclosed use of LLMs. Furthermore, LLMs can be useful in writing, reviewing and editing text, and promote equity in science. Others have argued that LLMs should be mentioned in the acknowledgments since they do not meet all the authorship criteria. We argue that naming LLMs as authors or mentioning them in the acknowledgments are both inappropriate forms of recognition because LLMs do not have free will and therefore cannot be held morally or legally responsible for what they do. Tools in general, and software in particular, are usually cited in-text, followed by being mentioned in the references. We provide suggestions to improve APA Style for referencing ChatGPT to specifically indicate the contributor who used LLMs (because interactions are stored on personal user accounts), the used version and model (because the same version could use different language models and generate dissimilar responses, e.g., ChatGPT May 12 Version GPT3.5 or GPT4), and the time of usage (because LLMs evolve fast and generate dissimilar responses over time). We recommend that researchers who use LLMs: (1) disclose their use in the introduction or methods section to transparently describe details such as used prompts and note which parts of the text are affected, (2) use in-text citations and references (to recognize their used applications and improve findability and indexing), and (3) record and submit their relevant interactions with LLMs as supplementary material or appendices. (shrink)

Ethics plays a key role in the normative analysis of the impacts of technology. We know that computers in general and the processing of data, the use of artificial intelligence, and the combination of computers and/or artificial intelligence with robotics are all associated with ethically relevant implications for individuals, groups, and society. In this article, we argue that while all technologies are ethically relevant, there is no need to create a separate ‘ethics of X’ or ‘X ethics’ for each and (...) every subtype of technology or technological property—e.g. computer ethics, AI ethics, data ethics, information ethics, robot ethics, and machine ethics. Specific technologies might have specific impacts, but we argue that they are often sufficiently covered and understood through already established higher-level domains of ethics. Furthermore, the proliferation of tech ethics is problematic because (a) the conceptual boundaries between the subfields are not well-defined, (b) it leads to a duplication of effort and constant reinventing the wheel, and (c) there is danger that participants overlook or ignore more fundamental ethical insights and truths. The key to avoiding such outcomes lies in a taking the discipline of ethics seriously, and we consequently begin with a brief description of what ethics is, before presenting the main forms of technology related ethics. Through this process, we develop a hierarchy of technology ethics, which can be used by developers and engineers, researchers, or regulators who seek an understanding of the ethical implications of technology. We close by deducing two principles for positioning ethical analysis which will, in combination with the hierarchy, promote the leveraging of existing knowledge and help us to avoid an exaggerated proliferation of tech ethics. (shrink)

In the original publication of this article, the Table 1 has been published in a low resolution. Now a larger version of Table 1 is published in this correction. The publisher apologizes for the error made during production.

The new wave of ‘foundation models’—general-purpose generative AI models, for production of text (e.g., ChatGPT) or images (e.g., MidJourney)—represent a dramatic advance in the state of the art for AI. But their use also introduces a range of new risks, which has prompted an ongoing conversation about possible regulatory mechanisms. Here we propose a specific principle that should be incorporated into legislation: that any organization developing a foundation model intended for public use must demonstrate a reliable detection mechanism for the (...) content it generates, as a condition of its public release. The detection mechanism should be made publicly available in a tool that allows users to query, for an arbitrary item of content, whether the item was generated (wholly or partly) by the model. In this paper, we argue that this requirement is technically feasible and would play an important role in reducing certain risks from new AI models in many domains. We also outline a number of options for the tool’s design, and summarize a number of points where further input from policymakers and researchers would be required. (shrink)

Patient online record access (ORA) is spreading worldwide, and in some countries, including Sweden, and the USA, access is advanced with patients obtaining rapid access to their full records. In the UK context, from 31 October 2023 as part of the new NHS England general practitioner (GP) contract it will be mandatory for GPs to offer ORA to patients aged 16 and older. Patients report many benefits from reading their clinical records including feeling more empowered, better understanding and remembering their (...) treatment plan, and greater awareness about medications including possible adverse effects. However, a variety of indirect evidence suggests these benefits are unlikely to accrue without supplementation from internet-based resources. Using such routes to augment interpretation of the data and notes housed in electronic health records, however, comes with trade-offs in terms of exposing sensitive patient information to internet corporations. Furthermore, increased work burdens on clinicians, including the unique demands of ORA, combined with the easy availability and capability of a new generation of large language model (LLM)-powered chatbots, create a perfect collision course for exposing sensitive patient information to private tech companies. This paper surveys how ORA intersects with internet associated privacy risks and offers a variety of multilevel suggestions for how these risks might be better mitigated. (shrink)