Prohibit public deployment of AI agents that are presented as, or indistinguishable from, actual human beings (living or dead) unless explicit provenance and consent standards are met. The policy would require registration, immutable provenance labels, and strict limits on recreating identifiable deceased people or fabricating new believable human identities.
— This frames a concrete regulatory option that addresses scams, grieving harms, political manipulation, and public‑trust collapse from synthetic humans.
Anil Seth
2026.10.07
100% relevant
Tavos's Griffin claim of passing a video Turing Test and the industry push for 'deadbots'—avatars trained on Holocaust survivors and deceased relatives—illustrate the technology and use cases this idea would regulate.
← Back to all ideas