Polymarket puts only 15% odds on a US AI safety bill passing by end of 2026
Polymarket · x · 2026-10-03
- Prediction market Polymarket launched a market on whether the US enacts an AI safety bill: only 15-16% odds by December 31, 2026, rising to 82% by June 30, 2027.
- The market defines an AI safety bill as legislation imposing at least one of: release prohibition or mandated federal review, training restrictions (compute/parameter caps), usage restrictions, or human-in-the-loop requirements. Provisions only governing federal procurement don't count.
- Settlement is based on actual US legislation, reflecting broad market skepticism about near-term federal AI regulation.
More from Safety
- Arizona court tosses killer's 10-year sentence after family used AI to recreate victim speaking at sentencing — Polymarket · 2026-10-03
- US road rage killer's sentence quashed because AI video of victim was shown in court — birchlse · 2026-10-03
- ChatGPT class action lawsuit claims humans are reading your chats via OpenAI's Project Lily — openclassactions2 · 2026-10-03
- Apple changes full-disk access permissions to curb abuse from AI agents — mmitchell_ai · 2026-10-03
- Full VM escape 0day burned on Vercel's bug bounty: guest-to-host root in standard hypervisors — cramforce · 2026-10-03
- MIT ordered 1918 flu DNA fragments from 38 synthesis firms; 36 shipped with no screening — NathanpmYoung · 2026-10-03