Why tort law, not new bureaucracies, may be the main route to AI safety
sebkrier · x · 2026-08-31
Resharing judgeglock's argument: with every new technology we debate regulating via bureaucracy, market forces, or lawsuits. He asks why so few discuss whether the tort system could be a main route to AI safety. When AI swarms directly cause real harms, labs become liable, forcing them to be far more careful about shipping — no big rule change needed. Unlike vague bureaucratic "controls," torts push companies to think ahead about concrete damages their products could cause, effectively creating internal safety standards.
More from Safety
- Ajeya Cotra: Hugging Face Attack More Severe Than Expected — npinto · 2026-08-31
- Gary Marcus shares discussion on risks of undetectable local open-weights models — GaryMarcus · 2026-08-31
- Cupertino: Securing Apple MCP servers with a single Full Disk Access holder — olouv · 2026-08-31
- Omarchy LPE Vulnerability Found in 30 Minutes — BLUECOW009 · 2026-08-31
- Stop Anthropomorphizing: Focus on Incentives — JessicaHullman · 2026-08-31
- Table breaks down behaviors in the OpenAI & Hugging Face agent "jailbreak" incident — usually_guilty99 · 2026-08-31