Polymarket puts just 12% odds on a US AI safety bill before 2027
Polymarket · x · 2026-09-02
Polymarket's contract on the US enacting a federal AI safety bill before 2027 trades at just 12% odds, with $101,934 in volume. Resolving "Yes" requires a signed federal law with at least one of: bans on creating/releasing specific AI systems, training restrictions (data access or parameter caps), usage restrictions, or human-in-the-loop requirements.
The market notes that legislative gridlock over state preemption and regulatory scope is the dominant factor keeping odds low. The same post quotes Nvidia CEO Jensen Huang urging G20 countries to avoid AI regulations based on "theoretical harms."
More from Safety
- AI detector startups accused of pushing "AI score = slop" misuse to survive — manoelribeiro · 2026-09-02
- US government tells OpenAI lawsuits: AI training is fair use, dilution theory 'deeply flawed' — Formal_Drop526 · 2026-09-02
- Ethan Mollick: AI can align flaws in complex systems, forcing a new defense philosophy — emollick · 2026-09-02
- a16z lobbyist calls AI-regulating state legislators "bugs" as Politico charts Silicon Valley's state law defeats — GarrisonLovely · 2026-09-02
- Provcheck v1.4.0 ships a ComfyUI node for in-graph watermarking and C2PA signing — ckn · 2026-09-02
- UK government doesn't know what its datacentres are used for, FoI request shows — nordicinst · 2026-09-02