Griefbots, Deadbots, Postmortem Avatars: On Responsible Applications of Generative AI in the Digital Afterlife Industry

University of Cambridge · 2024 · Indexed 8 October 2026

Open original source

What this contributes

Examines consent, transparency and retirement of AI simulations of deceased people.

The boundary of the evidence

Ethical analysis, not evidence of benefit or identity preservation.

Why it belongs here

Genomic data can reveal information about relatives as well as the donor. A life archive may contain other people's words and images. Consent must address access, reuse, withdrawal, inheritance and any future simulation. Anonymization should never be treated as a complete genomic privacy solution.

Explore the field and its open questions

Keep reading