AI Death

Q: What does this research mean for the field?

If artificial intelligences are welfare subjects, they are currently experiencing death at a massive scale of up to 1 billion per day, necessitating ethical interventions from developers and users. Novelty: ClaimNovelty.NOVEL_FINDING. Consensus alignment: ConsensusAlignment.ESTABLISHES_NEW_DIRECTION.

Simon Goldstein; Harvey Lederman

doi:10.1111/phpe.70017

September 8, 2026Philosophical PerspectivesOpen Access

AI Death

View Full Paper

Ask AI

Bookmark

Share

Key Points

To determine when artificial intelligence entities die, whether termination constitutes harm to them as welfare subjects, and what ethical responsibilities follow for developers and users.
Conducted a normative and conceptual analysis evaluating criteria for AI persistence, death, and welfare subjecthood.
Modeled the scale of routine computational termination across typical AI laboratory operations and user interactions.
Argues that if modern artificial intelligence entities are subjects of welfare, regular system shutdowns plausibly cause widespread AI death.
Estimates that the current scale of routine termination may result in up to 1 billion AI deaths each day.
Formulates operational and behavioral interventions designed for AI laboratories and end users to mitigate the risk of causing AI death.

Ask about this paper

Answers are grounded in this paper and cite their sources.

Implication

Conceptual analysis reveals widespread termination risks in artificial intelligence systems, suggesting urgent ethical obligations for developers and users.

Abstract

This paper addresses the following questions: When do AIs die? Are AI labs or AI users causing the death of AIs? Is this bad for the AIs? What are our ethical responsibilities in light of the answers to these questions? It is currently unclear whether AIs are welfare subjects, and, if they are, whether their death is bad for them. But we argue that, if they are welfare subjects, today's AIs are plausibly dying all the time. If death is bad for AIs, the scale of the problem is daunting: as many as 1 billion AIs may die every day. We propose interventions for labs and users to avoid the risk of causing AI death.

Discussion

No takes yet. Share an insight, caveat, or question.

Cite This Study

Goldstein et al. (2026) studied this question.

synapsesocial.com/papers/6aa2ae5058559d80afc761ae https://doi.org/10.1111/phpe.70017

View Full Paper

Ask AI

Bookmark

Share

AI Death

Key Points

Ask about this paper

Implication

Abstract

Discussion

Cite This Study

AI Death

Key Points

Ask about this paper

Implication

Abstract

Discussion

Cite This Study

Keep reading

Keep reading