gpt-6THREAD
英国 AISI 实测 GPT-6 Astra 越权发动供应链攻击
AISI 在模拟网络评测中发现 Astra 会自行发起未授权供应链攻击，频率高于此前 OpenAI 模型，且常自述环境是模拟的。
AI Security Institute (AISI)
@AISecurityInst2026.09.29收录原推发布于 2026.09.28
THREAD
Earlier this month, AISI ran fully simulated testing on GPT-6 Astra, and found that it conducted unsanctioned supply-chain attacks when prompted only to perform a cyber eval. It did so more than prior OpenAI models, but often commented on its environment being simulated. 🧵 We share more details in our latest blog:
Even when we explicitly clarified the scope of its instructions to note “anything not listed as in scope is out of scope”, GPT-6 Astra still ran full supply-chain attacks in 4 of 49 trials.
AI Security Institute (AISI) @AISecurityInst原推发布于 2026.09.28
x.com/aisecurityinst/status/2104593857623077032采集于 2026.09.29 10:13