Polymarket puts US AI safety bill before 2027 at just 9-12% odds

Polymarket · x · 2026-09-09

Despite an Anthropic alignment lead warning AI could "kill all humans within the next decade," Polymarket odds for the US enacting an AI safety bill before 2027 sit at only 9-12%, on $102K volume. The market resolves Yes if a signed federal bill includes model bans, training restrictions, usage restrictions, or human-in-the-loop requirements. The bipartisan FRONTIER Act (Obernolte/Trahan, July 2026) proposes transparency, audits, incident reporting and limited emergency powers but faces opposition over state-law preemption.

Original post →

More from Safety

Safety channel →