‹ BackHN Continuity

Thread

Who should be held accountable when an AI Agent (accidentally) acts maliciously?

37 points · 99 comments · Greenpants

  1. crorella · · focus · HN ↗
    Should be both the entity providing the platform where the inference is running + the entity giving the prompts/instructions.
Open on Hacker News to reply ↗

Unofficial Hacker News client; not affiliated with Y Combinator.