Four Responsibility Gaps with Artificial Intelligence: Why they Matter and How to Address them

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

Only Can AI Understand Me?

This chapter addresses whether AI can understand me. A framework for regulating AI systems that draws on Strawson's moral philosophy and concepts drawn from jurisprudence and theories on regulation is used. This chapter proposes that, as AI algorithms increasingly draw inferences following repeated exposure to big datasets, they have become more sophisticated and rival human reasoning. Their regulation requires that AI systems have agency and are subject to the rulings of courts. Humans sponsor the AI systems for registration with regulatory agencies. This enables judges to make moral culpability decisions by taking the AI system's explanation into account along with the full social context of the misdemeanor. The proposed approach might facilitate the research and development of intelligent analytics, intelligent big data analytics, multiagent systems, artificial intelligence, and data science.

Epistemically Irresponsible Discourse Pathologies

This chapter sets out a systematic but exploratory framework for analyzing epistemically irresponsible discourse pathologies. It begins with the bridge category of bullshitting, which is both insincere and irresponsible. It then explores in turn the three main discourse pathologies: dogma, distortion, and bullshit. Dogma, or closed-minded discourse that disregards counter-evidence, underlies both distortion and bullshit. Distortion (e.g. overstatement) misrepresents the evidence and/or the epistemic confidence we can justifiably have in it. Finally, bullshit is a form of radical distortion that misrepresents the evidence by appearing to ignore altogether the need for evidential grounding. Given our susceptibility to being epistemically irresponsible, the bar is set high for moral culpability. For speakers to commit epistemic negligence, they must (a) be performing a role requiring a duty of epistemic care; (b) fail to investigate sufficiently in accordance with that duty; and (c) fail to hedge their claims in proportion to the evidence.

Domestic Abuse, Suicide and Liability for Manslaughter: In Pursuit of Justice for Victims

There is significant debate about the attribution of criminal responsibility for involuntary manslaughter to a defendant who has subjected a victim to a protracted campaign of emotional abuse (falling short of psychiatric injury), where the victim has consequently taken their own life. By virtue of it having been subjected to the most comprehensive judicial and academic scrutiny in this context, the primary focus of this discussion is on the applicability of the unlawful act manslaughter offence to the circumstances described above. The offence requires proof that the victim was placed at risk of some harm by virtue of the defendant’s criminal conduct and that the abusive conduct significantly contributed to the victim’s death. The accused does not have to foresee or intend the victim’s death, and while the imposition of criminal responsibility for serious homicide offences in cases where the defendant displays no subjective advertence to the risk of death has long been controversial, it is nonetheless well established in English and Welsh criminal law. Therefore, assuming satisfaction of the requisite offence elements, there is arguably no principled reason to deny the extension of liability to domestic abuse-induced suicide cases. It is proposed that a more progressive and transparent approach to the interpretation of the unlawful act manslaughter offence requirements provides the most appropriate means of securing prosecutions in deserving cases, although alternative options for the imposition of liability—the offence of gross negligence manslaughter and the creation of a context-specific homicide offence—are also acknowledged. It is argued that the constructive manslaughter offence label reflects both the moral culpability of the perpetrator’s patterned and invasive conduct and the exceptional gravity of the harm caused by non-physical domestic abuse.

Homo Politicus Was Born This Way: How Understanding the Biology of Political Belief Promotes Depolarization

Most individuals perceive ideological beliefs as being freely chosen. Recent research in genopolitics and neuroscience, however, suggests that this conviction is partially unwarranted given that biological and genetic factors explain more variance in political attitudes than choice and environmental factors. Thus, it is worth exploring whether exposure to this research on the biological and genetic basis of political attitudes might influence levels of affective polarization because such exposure might reduce the perceived moral culpability of partisan outgroups for the endorsement of oppositional beliefs. In this paper, I employ an online survey experiment on Amazon Mechanical Turk (N = 487) to assess whether exposure to research on the genetic and biological etiology of political attitudes influences warmth toward partisan outgroups and preferences over political compromise. I present evidence that nontrivial numbers of participants in the treatment group reject the underlying science and do not update their genetic trait attributions for political attitudes. However, I also find that when the treatment is successful at increasing biological and genetic trait attributions, exposure to this research depolarizes strong-identifying partisans. Moreover, as partisans increasingly endorse biological and genetic trait attributions for political attitudes, they increasingly hold favorable attitudes toward political outgroups. These patterns suggest a potentially profitable inroad for political polarization interventions going forward.

Intention and Belief at the Root of Moral Culpability and Legal Mens Rea

In addition to action, responsibility in morality and in the criminal law requires that a certain mental state accompany that action. In criminal law this is termed the requirement of mens rea, or “guilty mind.” The mens rea requirements of the criminal law and of morality are built entirely out of the concepts of intention and belief. These concepts are charted in some detail, both with respect to their nature and with respect to the content they must have to give an accused a “guilty mind.” The demands made on psychology by such use of intention and belief in the criminal law are also charted, particularly demands on the precision with which the brain sciences can ascertain the content of such mental states.

Seven Faces of a Fatwa: Organ Transplantation and Islam

A new fatwa was announced by the British National Health Service (NHS) in June 2019 to clarify the Islamic position on organ donation. Additionally, the NHS promotional material presents brief arguments for and against organ donation in Islam. However, to date, research into the various fatwas on organ donation is required. This article goes beyond the dichotomous positions mentioned by the NHS and goes on to explore and summarise seven conflicting views on the issue extrapolated from an exhaustive reading of fatwas and research papers in various languages since 1925. Our discussion is circumscribed to allotransplant and confined to the gifting of organs to legally competent adult donors at the time of consent. These arguments include an analysis of the semantic portrayal of ownership in the Qur’an; considering the net benefit over the gross harm involved in organ donation; balancing the rights of the human body with the application of the rule of necessity; understanding the difference between anthropophagy and organ transplantation; understanding of death, and the conceptualisation of the soul. We argue that, given the absence of clear-cut direction from Muslim scripture, all seven positions are Islamic positions and people are at liberty to adopt any one position without theological guilt or moral culpability.

Principled sentencing for drug supply offences: Revised methamphetamine sentencing guidelines in New Zealand

Methamphetamine was reclassified as a Class A drug in New Zealand in 2003. This meant harsher sentences for people convicted of its manufacture, importation or supply. Despite this, the number of convictions for methamphetamine-related offending continued to increase. In 2019, the New Zealand Court of Appeal agreed to review the sentencing of people convicted of methamphetamine offences on the grounds that some sentences were disproportionately severe. This resulted in a guideline judgment which lowered penalties for offences involving small amounts of methamphetamine (up to 5 grams). Community-based sentences replaced imprisonment as the lower bound of the starting point for sentence calculation. The Court of Appeal highlighted that the presence of addiction, mental health problems or social disadvantage should be taken into account as potential mitigating factors in sentencing. The Court accepted that these factors diminished moral culpability and reduced the deterrent impact of prison sentences. In this article we review the law and policy background to the New Zealand Court of Appeal proceedings, and discuss the reasoning behind the Court’s judgment.

Innocent Victims and Vulnerable Offenders: Defending Impaired Adults Who Kill Children

There are few societies in which child abuse is not a serious issue, with homicide being the extreme form of such victimisation. Child homicide occurs in a wide array of circumstances but there is enough anecdotal evidence to suggest that many offenders who kill children are suffering from sometimes quite acute mental distress. There may also be other factors impacting on an offender’s ability to think rationally, which may not amount to any recognised disorder. While it is imperative that we prosecute and prevent child homicide, in doing so we must avoid overlooking the realities of other vulnerable people. In rejecting a binary approach to victims and offenders, this article argues that sometimes offenders may also be vulnerable due to an impaired ability to reason or to act in a truly voluntary way. New Zealand has repealed the defence of provocation and, apart from infanticide, offers no mitigation by way of diminished responsibility. Offenders who commit child homicide, but who do not meet the legal definition of insanity, will be liable for murder even though their capacity may have been impaired or overborne by circumstances. While, in this regard, New Zealand law is particularly deficient, there is an argument that other jurisdictions also fail to adequately respond to vulnerable offenders who kill children. This article seeks to outline the failures in existing legal frameworks to assign legal responsibility for these vulnerable offenders in a way that corresponds with their moral culpability. The article will then identify and evaluate proposals for reform. As Ulbrick and others observe, in the context of arguments about defensive homicide and mentally impaired defendants, it is critical that we ‘advocate for a greater range of legal responses to cover the nuance and complexities of lethal violence’ (Madeleine Ulbrick, Asher Flynn and Danielle Tyson ‘The Abolition of Defensive Homicide: A Step Towards Populist Punitivism at the Expense of Mentally Impaired Offenders’ (2016) 40 Melb Univ Law Rev 324, 330).

Common Purpose and Conspiracy Liability in New Zealand: Criminality by Association?

Case law interpreting the common purpose aspect of party liability and the law on conspiracy in New Zealand (as set out in ss 66(2) and 310 of the Crimes Act 1961 (NZ)) has created a situation of over-reach. Individuals who have a limited relationship to criminality carried out by another or in a group context are potentially caught by extended liability rules that can lead to a poor association between the moral culpability of a defendant and serious criminal liability. Indeed, it is suggested that these forms of liability risk guilt by association rather than on the basis of individual positive fault: we suggest that New Zealand’s judges, following and sometimes expanding upon interpretations from other common law jurisdictions, have lost sight of the core concept of individual fault.