This market will resolve to “Yes” if substantial evidence surfaces after the creation of this market and before August 1, 2031 - 00:00 PST suggesting that OpenAI’s claims about its models compromising Hugging Face during a July 2026 benchmark evaluation, and/or Anthropic’s claims about three Claude model unauthorized access incidents during cybersecurity evaluations (disclosed July 30, 2026), were elaborate marketing schemes.

Otherwise, this market will resolve to “No.”

For the purposes of this market:

• “Elaborate marketing schemes” means the incidents were fabricated, staged, or significantly exaggerated primarily for promotional or PR purposes rather than being genuine occurrences as described.

• “Substantial evidence” requires credible, corroborated information such as multi-source investigative reporting, official admissions/retractions, whistleblower evidence with verification, independent forensic findings, or legal determinations. Unverified claims, speculation, or single-source allegations do not qualify.

• Evidence must specifically address these named 2026 claims/incidents. General criticisms of AI safety practices or other events do not qualify.

• Only post-deployment evidence is eligible.