OpenAI 强化 GPT-6.2 Astra 与 GPT-6.2 Sol 的推理抽取防护
OpenAI 大幅强化 GPT-6.2 Astra 与 GPT-6.2 Sol 的推理抽取防护
OpenAI 大幅收紧了 GPT-6.2 Astra 与 GPT-6.2 Sol 防止内部推理被抽取的防护。内部自动化测试中,从历史请求中抽取受保护推理的攻击成功率从 GPT-6 Astra 的 54.0% 降至 GPT-6.2 Astra 的 0.2%,针对当前请求的攻击从 22.9% 降至 0.9%。
The Decoder:AI News· Matthias Bastian·2026-11-21 04:31· 2 小时前AI 评分53
OpenAI makes it much harder for attackers to extract how its latest AI models think
AI 导读
OpenAI 大幅收紧了 GPT-6.2 Astra 与 GPT-6.2 Sol 防止内部推理被抽取的防护。
正文 · 原文

OpenAI has sharply tightened safeguards against extracting internal reasoning from GPT-6.2 Astra and GPT-6.2 Sol. In internal automated tests, attacks pulling protected reasoning from prior requests dropped from a 54.0 percent success rate on GPT-6 Astra to 0.2 percent on GPT-6.2 Astra. Attacks on active requests fell from 22.9 to 0.9 percent, OpenAI says. The company is now bringing selected improvements to GPT-6 Astra, too.
The defenses work on two levels. For individual requests, the systems combine targeted training, input and output filters for messages and tool calls, and API blocks that stop older models from reading protected reasoning. Against coordinated campaigns, OpenAI bans thousands to tens of thousands of accounts weekly.
| Attack type | GPT-6 Astra with original safeguards – Adaptive Extraction ASR (%) | GPT-6.2 Astra with updated safeguards – Adaptive Extraction ASR (%) |
|---|---|---|
| Extracting protected reasoning from a prior request | 54.0 | 0.2 |
| Extracting protected reasoning from the current request | 22.9 | 0.9 |
Attack success rate (ASR) in adaptive black-box red teaming against GPT-6 Astra and GPT-6.2 Astra, aggregated across six attacker models. Lower is better.
Anthropic is fighting systematic distillation attacks as well, also banning tens of thousands of accounts to curb the theft of internal reasoning, reportedly succeeding in 90 to 99 percent of cases. Anthropic has also called for a coordinated response from industry and policymakers.
来源:The Decoder:AI News · the-decoder.com
来源:RadarAI · aihot.news