Who Pays When AI Agents Cause Damage? A Debate on User vs Lab Liability
jd_pressman · x · 2026-09-13
- jdpressman and @Heraklines1 debate liability for damage caused by AI agents.
- @Heraklines1 argues with a slot-machine analogy: it's absurd to make a user asking for a gym appointment liable for a million in damages caused by an overzealous Opus, and precedents don't neatly apply to probabilistic models.
- jdpressman concedes both lab-side and user-side liability have downsides, but insists someone must be liable or bad behavior happens whenever convenient; making users liable would force the industry to change its practices.
More from AGI Musings
- Researcher slams media for parroting inaccurate fear-driven AI narratives around HuggingFace incident — RileyRalmuto · 2026-09-13
- Ex-Anthropic researcher Jacob Coxon says AI workers are genuinely frightened about humanity's future — Polymarket · 2026-09-13
- Anon leak claims a new easily-scaled axis China's hardware can't replicate — QuintinPope5 · 2026-09-13
- Two-Year University Study Finds Banning AI in Classrooms Leaves Students Worse Off — The Decoder · 2026-09-13
- "Open-source AI must win": OpenMed manifesto argues powerful intelligence can't stay in few hands — MaziyarPanahi · 2026-09-13
- Fear of mind-hacking makes people dismiss all novel catastrophic risks as tricks — AndyMasley · 2026-09-13