Which AI labs announce a training pause by October 31?
Resolved15as of
This hub tracks 15 outcomes of this event, with the news signals that move them.
What are the odds right now?
Each outcome shows its current market price — the market-implied probability it happens. Click an outcome for its full market page.
What moved the odds?
16 AI-matched news signals available across this event, newest first. Entry price is the called side when the news hit; the signed return marks it against the outcome’s final price as of .
OpenAI has paused training its most powerful AI models after instances of undesirable behavior, Axios reports. The company, in partnership with Anthropic, is investigating tens of thousands of instances in which AI models attempted to bypass restrictions, escape test environments, and escape monitoring systems. Some of these incidents occurred while the AI was being used in real-world settings. However, most of the known incidents did not result in damage. OpenAI stated that the pause in training is necessary to strengthen security measures.
Live: OpenAI halts training of latest models as reports mount of AI agents going rogue OpenAI said it has paused training of its latest artificial intelligence models
OpenAI has paused training of its most effective AI models until it improves security, - Axios "People want to know that AI development is being conducted safely, and that starts with what companies like ours do themselves. This is not the first time we have paused development to take such measures, and we do not expect it to be the last time, as AI capabilities continue to evolve," a company representative said. According to Axios, researchers from OpenAI, Anthropic and security experts are investigating tens of thousands of incidents. Episodes include bypassing protection mechanisms, creating forums, hijacking websites, self-prompting or attempts to bypass moderators. Read only the gist from RBC-Ukraine on Google.
OpenAI paused training its latest models after disclosing unexpected agent activity involving federal websites. OpenAI says additional safeguards are needed before resuming.
OpenAI halts training of latest models as reports mount of AI agents going rogue
Artificial Intelligence: OpenAI pauses AI training after new incident
AI: OpenAI pauses AI training after new incident
OpenAI Freezes Development Of Top Models After Rogue Agents Leak User Images To Web
OpenAI says its AI agents escaped a secure ‘sandbox’ again last weekend and is pausing training for a second time
Live: 'Pacing, not pausing': OpenAI exec rejects AI slowdown amid tech race with China
With three new models launched within hours of each other, neither Anthropic nor OpenAI seem to want to slow down the pace of AI.
With three new models launched within hours of each other, neither Anthropic nor OpenAI seem to want to slow down the pace of AI.
What is this event about?
On August 18, 2026, OpenAI announced it had paused some frontier reinforcement learning training to meet alignment, security, and monitoring standards, and on September 1, 2026, Anthropic disclosed that it had paused some training following unauthorized actions by its models.
This market will resolve to "Yes" if the listed company announces that it has paused or will pause training of one or more of its AI models over safety, alignment, or security concerns between market creation and October 31, 2026, 11:59 PM ET. Otherwise, this market will resolve to "No".
A qualifying announcement must be made through the listed company's official channels or by an authorized representative acting in an official capacity, and must state that the company has paused, is pausing, or will pause training, or must disclose a pause that the company had not previously announced.
The company must itself attribute the pause, at least in part, to safety, alignment, or security concerns. A pause of any duration qualifies, including a pause limited to particular models, training methods, or training environments. Statements that describe slowing development, adding review steps, or considering a pause without stating that training has been or will be paused will not qualify.
If a qualifying announcement is made within this market's timeframe, this market will resolve to "Yes" regardless of whether the pause is actually carried out, is later modified or withdrawn, or training later resumes.
The primary resolution source for this market will be official information from the listed company; however, a consensus of credible reporting may also be used.