What would qualify an artificial intelligence for moral standing?

Ladak, A.

AI Ethics 4, 213–228 (2024).

https://doi.org/10.1007/s43681-023-00260-1

Abstract

What criteria must an artificial intelligence (AI) satisfy to qualify for moral standing? My starting point is that sentient AIs should qualify for moral standing. But future AIs may have unusual combinations of cognitive capacities, such as a high level of cognitive sophistication without sentience. This raises the question of whether sentience is a necessary criterion for moral standing, or merely sufficient. After reviewing nine criteria that have been proposed in the literature, I suggest that there is a strong case for thinking that some non-sentient AIs, such as those that are conscious and have non-valenced preferences and goals, and those that are non-conscious and have sufficiently cognitively complex preferences and goals, should qualify for moral standing. After responding to some challenges, I tentatively argue that taking into account uncertainty about which criteria an entity must satisfy to qualify for moral standing, and strategic considerations such as how such decisions will affect humans and other sentient entities, further supports granting moral standing to some non-sentient AIs. I highlight three implications: that the issue of AI moral standing may be more important, in terms of scale and urgency, than if either sentience or consciousness is necessary; that researchers working on policies designed to be inclusive of sentient AIs should broaden their scope to include all AIs with morally relevant interests; and even those who think AIs cannot be sentient or conscious should take the issue seriously. However, much uncertainty about these considerations remains, making this an important topic for future research.

Here are some thoughts:

This article explores the criteria for artificial intelligence (AI) to be considered morally significant. While sentience is often seen as a requirement, the author argues that some non-sentient AIs might qualify too. The paper examines different viewpoints and proposes that AIs with complex goals or consciousness, even without sentience, could be morally relevant. This perspective suggests the issue might be broader and more pressing than previously thought. It also highlights the need for AI policies to consider a wider range of AI and for those skeptical of AI sentience to still acknowledge the moral questions it raises. Further research is needed due to the remaining uncertainties.

Autonomy and the Folk Concept of Valid Consent

Demaree-Cotton, J., & Sommers, R. 

(2021, August 17). 

https://doi.org/10.31234/osf.io/p4w8g

Abstract

Consent governs innumerable everyday social interactions, including sex, medical exams, the use of property, and economic transactions. Yet little is known about how ordinary people reason about the validity of consent. Across the domains of sex, medicine, and police entry, Study 1 showed that when agents lack autonomous decision-making capacities, participants are less likely to view their consent as valid; however, failing to exercise this capacity and deciding in a nonautonomous way did not reduce consent judgments. Study 2 found that specific and concrete incapacities reduced judgments of valid consent, but failing to exercise these specific capacities did not, even when the consenter makes an irrational and inauthentic decision. Finally, Study 3 showed that the effect of autonomy on judgments of valid consent carries important downstream consequences for moral reasoning about the rights and obligations of third parties, even when the consented-to action is morally wrong. Overall, these findings suggest that laypeople embrace a normative, domain-general concept of valid consent that depends consistently on the possession of autonomous capacities, but not on the exercise of these capacities. Autonomous decisions and autonomous capacities thus play divergent roles in moral reasoning about consent interactions: while the former appears relevant for assessing the wrongfulness of consented-to acts, the latter plays a role in whether consent is regarded as authoritative and therefore as transforming moral rights.

Conclusion 

Before these studies, it remained an open possibility that “valid consent” as a rich and normatively complex force existed only as a technical concept used in philosophical, legal and academic domains. We found, however, that the folk concept of consent involves normative distinctions between valid and invalid consent that are sensitive to the consenter’s autonomy, even if the linguistic utterance of “yes” is held constant, and that this concept plays an important role in moral reasoning. 

Specifically, the studies presented here examined the relationship between autonomy and intuitive judgments of valid consent in several domains: medical procedures, sexual relations, police searches, and agreements between buyers and sellers.  Across scenarios, we found that judgments of valid consent carried a specific relationship to autonomy: whether an agent possesses the mental capacity to make decisions in an autonomous way has a consistent impact on whether their consent is regarded as valid, and thus whether it was regarded as morally transformative of the rights and obligations of the consenter and of third parties.  Yet, whether the agent in fact makes their decision in an autonomous, rational way—based on their own authentic values and what is right for them—has little impact on perceptions of consent or associated rights, although it has relevance for whether the consent-obtainer is acting wrongly.  Autonomy thus has a subtle role in the ordinary reasoning about morally transformative consent, where consent given by an agent with autonomous capacities has a distinctive role in downstream moral reasoning.

The research is here.