Analytics
Vultax does not execute orders. These controls open the official Polymarket market or route back into the Vultax predictions terminal.
“Anthropic reports another AI sandbox escape by October 15?” is a tech prediction market on Polymarket with two tradeable outcomes, Yes and No. It belongs to the event “Anthropic reports another AI sandbox escape by...?”. The Yes price is the market's implied probability that the question resolves Yes. Quotes can change as orders are updated or filled; a quoted price does not guarantee an available fill at that price.
Trading is open until Oct 16, 2026 UTC. As of Sep 30, 9:53 PM UTC, Yes trades at 13.5¢, a 13.5% implied chance, $861.3 traded in the last 24 hours, $83.29 resting in the order book, a 11.0% bid-ask spread.
Vultax observes this market from Polymarket's public order book and data API and keeps its own record: 15 daily price points from Sep 16, 2026 UTC to Sep 30, 2026 UTC (Yes between 8.5% and 31.0%); 14 loaded holder wallets, the largest ten holding 99.0% of the loaded shares; 40 recent public fills from 22 wallets. Each figure carries its observation time; nothing here is a return forecast.
Public Polymarket market/event context is available; Vultax source-pack coverage is tracked separately.
Current public quote is shown when Vultax depth snapshots are unavailable.
Nine stops, one per figure on this page. Each stop says what the number is, why it matters, and what it reads right now for this market.
Each outcome trades as a share that pays $1 if it wins and nothing if it loses. Prices are quoted in cents, and Yes plus No add up to about one dollar.
Daily closes of the Yes price, Polymarket CLOB · 15 points · 16 Sep 2026 – 30 Sep 2026 · axis zoomed to 0%–40%
| Observed | Yes price |
|---|---|
| 16 Sep 2026 | 31.0% |
| 30 Sep 2026 | 17.5% |
| Trader | Outcome | Shares |
|---|---|---|
| Lexington | No | 376.1 |
| ENORMOUZ | Yes | 284.85 |
| SuperTests | Yes |
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 Anthropic publicly discloses an 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 Anthropic 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 Anthropic'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 Anthropic confirms the incident. The primary resolution source for this market will be official information from Anthropic; however, a consensus of credible reporting may also be used.
Official listing: polymarket.com/event/anthropic-reports-another-ai-sandbox-escape-by
The Yes price read as a chance. A share at 21.5¢ implies 21.5%, but fees, available depth and who participates can affect that price. A market price is not a calibrated forecast or a verdict.
Dollar value of shares that changed hands in the last 24 hours. High volume means the price is being tested by real money; low volume means a few trades can move it.
Reported dollars resting in the order book. The available size on your side, at the prices you would accept, determines what can fill. A large total alone does not guarantee an execution near the quoted price.
The gap between the best bid and the best ask. A tight spread makes entry and exit cheap; a wide spread is a cost paid on every round trip.
The wallets holding the most shares on each side, and how much of the loaded sample the largest ten control. Concentration shows whether a market is a crowd or a handful of large positions.
The public fills, newest first: which wallet bought or sold which outcome, at what price, and when. It is the market's own diary, and every wallet on it links to a trader profile.
Observed closes of the Yes price over the market's life. The range shows how sure or split the crowd has been; the move shows which way it has been changing its mind.
Trading stops at the close time. The outcome is then settled by the resolution source named in the rules. A closed market is not resolved until the venue reports it, so the two states are shown separately.
| 43 |
| Started-with-20-USD | Yes | 31 |
| AJSV | Yes | 22.53 |
| FrogCharts | No | 21 |
| romanew-1-new | Yes | 11.25 |
| TPredictions | No | 10 |
| TimPenis | Yes | 7.14 |
| Exemplary-Decisionmaking | Yes | 5 |
Every wallet's public record, and 80,000 ranked by lifetime P&L: Polymarket leaderboard
