OpenAI says GPT-6.2 Astra safeguards cut reasoning-extraction attack success to 0.2%
OpenAI announced on November 20, 2026 stronger defenses against adversarial distillation. It reports that adaptive attacks extracting protected reasoning from an earlier request succeed 0.2% of the time on GPT-6.2 Astra with updated safeguards, versus 54.0% on GPT-6 Astra with original safeguards.
OpenAI on November 20, 2026 announced stronger defenses against adversarial distillation, reporting that the average success rate for adaptive attacks extracting protected reasoning from an earlier request fell from 54.0% on GPT-6 Astra with original safeguards to 0.2% on GPT-6.2 Astra with updated safeguards.
The reported numbers
For extraction from the current request, the corresponding average fell from 22.9% to 0.9%, OpenAI reports.
| Attack target | GPT-6 Astra (original safeguards) | GPT-6.2 Astra (updated safeguards) |
|---|---|---|
| Protected reasoning from an earlier request | 54.0% | 0.2% |
| Protected reasoning from the current request | 22.9% | 0.9% |
According to OpenAI, the figures come from an adaptive black-box evaluation that aggregates results across six attackers. These results are OpenAI's own and have not been independently verified.
Context window, pricing and benchmark scores for GPT-6.2 Astra are not yet disclosed.
The announcement was covered by Unite.AI.
What this means
Adversarial distillation is the practice of using a model's outputs to train another model, and the reasoning traces are the protected content in this evaluation. OpenAI's figures show the updated safeguards on GPT-6.2 Astra performing far below the original safeguards on GPT-6 Astra in its own tests, for both earlier-request and current-request extraction. Because the evaluation is adaptive, the attackers adjust their approach rather than relying on fixed prompts. Developers who depend on reasoning protection should note that the numbers are OpenAI-reported averages across six attackers, not an independent audit.
Related Articles
OpenAI outlines four-part approach to combat adversarial distillation
OpenAI describes a layered approach to adversarial distillation. It combines model training, safeguards, detection, and enforcement to protect against capability extraction.
OpenAI tests ChatGPT dots with a small group of Plus users, rolls out to all Business Standard
OpenAI has started testing ChatGPT dots with a small number of ChatGPT Plus subscribers and is rolling it out to all ChatGPT Business Standard users. Dots launched in September for Pro, Business Premium and Enterprise plans. Usage allowances for the new tiers have not yet been clarified.
OpenAI report: median researcher's agent usage would cost $2,400 per day at API list prices in October
OpenAI's second internal research acceleration report says the median researcher would have consumed $2,400 of tokens per day at API list prices in October, up 62 percent from September. Agent success rates on longer tasks have risen, but the systems still fail at complex engineering work and need close human oversight.
OpenAI models optimized circuits used in Jalapeño silicon and improved RISC-V core designs
OpenAI reports that its models optimized circuits used in Jalapeño silicon and improved RISC-V core designs. The company presents the work as progress in AI-assisted chip design with GPT-6 Astra.
Comments
Loading...