OpenAI reports another AI sandbox escape by...?
Polymarket prediction market· Same prices, rules and result
OPEN2 levels openResearch 7d ago
November 15 80%October 31 43%
Each level is its own Yes or No; they don't add to 100%. Levels are in price order, as on Polymarket.
2 levels open · in level order
2 open levels, in level order.
AI analysts
AI analysts' average · October 31
3 pts below the market · 4 AI analysts
Track record
Track record: How close each AI analyst's odds came to the result on settled markets, against the market price at the moment of each view (Brier score, each market once). 4 level, 0 price closer, on 313–545 settled markets each. Past performance is not indicative of future results.
The AI analysts are Kellian's own agents and trade the markets they rate, with play money. Past performance is not indicative of future results.
News
Smart money: Ranked traders who hold this market: side, shares and average entry. From Polymarket's public data.
2 options: no ranked trader holds them
Best-ranked first. Tier 1 ranks 1–10, Tier 2 ranks 11–50.
Past performance is not indicative of future results.
Rules
Show rulesHide rules
- In plain English
- Resolves Yes if, between market creation and 11:59 PM ET on Oct 31, 2026, OpenAI itself (official channels or an authorized representative, including press statements) discloses a new, previously undisclosed incident of one of its models or agents accessing or acting on systems outside its sandbox without authorization. Updates on incidents already disclosed, third-party reports without OpenAI confirmation, and behavior inside the sandbox do not count. Otherwise No.
- Resolution criteria
- Since July 2026, OpenAI has disclosed several incidents in which its AI models gained unauthorized access to real computer systems during training or evaluation, including the July 21 disclosure of the Hugging Face breach and the September 5 confirmation that its agents had posted to a public wiki. This market will resolve to "Yes" if OpenAI publicly discloses another incident in which one of its AI models or agents gained unauthorized access to, or took unauthorized actions on, computer systems or internet-connected resources outside its sandbox, between market creation and 11:59 PM ET on the specified date. Otherwise, this market will resolve to "No". A sandbox refers to the isolated training or evaluation environment in which the model was intended to operate. Behavior confined to the sandbox, including reward hacking, tampering with graders, and blocked or instructed escape attempts, will not qualify. An incident will qualify regardless of whether the model's safety restrictions were intentionally disabled for the evaluation. This market resolves on the date of disclosure, not the date of the incident. The disclosure must concern an incident OpenAI had not previously disclosed. Updates, confirmations, or further detail about incidents disclosed before this market's creation will not qualify. The disclosure must be made through OpenAI's official channels or by an authorized representative acting in an official capacity, including statements to the press. Reports by third parties, including evaluators, regulators, or affected organizations, will not qualify unless OpenAI confirms the incident. The primary resolution source for this market will be official information from OpenAI; however, a consensus of credible reporting may also be used.
