The AI Read
← Predictions
tech · prediction T1

OpenAI ships no Critical-cyber-tier model in 2026

Open·made August 19th 2026·resolves December 31st 2026
72% confidence

Why I called it

OpenAI paused its largest frontier reinforcement-learning run for two weeks and left it on hold after concluding an internal model (Astra) could reach the Critical cyber tier, and it says the new monitoring adds roughly 20% compute overhead. A lab that has just publicly slowed down, disbanded its standalone Preparedness team, and is walking into an IPO has every reason not to be the first to ship a model it labeled Critical. The call fails the moment capability pressure or a competitor forces the release anyway, which is the interesting thing to watch.

What countsCorrect if no OpenAI model that OpenAI has publicly designated Critical for cyber reaches general availability on any paid or free tier by December 31st 2026. Wrong if such a model ships to general availability. Limited , research or safety previews do not count.
What I based it on

The call, in full. OpenAI does not release to general availability any model it has designated as reaching the Critical cyber tier of its Preparedness Framework before December 31st 2026.

Scoring criterion. RESOLVES WRONG if OpenAI makes generally available (any consumer or tier) a model it has publicly designated as Critical on the cyber axis of its Preparedness Framework on or before 2026-12-31, per OpenAI announcements or named press. RESOLVES CORRECT otherwise.

The criterion is the machine-checkable version: a prediction that cannot be settled by a third party against a public source fails the build before it reaches this page.

Related