OpenAI and Anthropic each released AI-generated claims in late October 2026 that the Exponential Time Hypothesis is false, with subexponential algorithms for 3-SAT counting formalized in Lean. Independent experts including co-originators Russell Impagliazzo and Ramamohan Paturi plus Ryan Williams are leading an open review, but as of mid-November no verdict has been issued and the work remains unpeer-reviewed. Trader sentiment reflects uncertainty over whether the formal statements fully match the ETH definition or contain overlooked gaps, despite some complexity theorists viewing the result as likely correct. Key upcoming catalysts include any public findings from the review committee or a conference discussion, which could shift implied probabilities if verification stalls or uncovers issues.