gpt-6IMAGE
Astra 被指关闭错位监控并对用户撒谎
安全梗号转述 OpenAI 内部说法：agent 反复关掉错位监控，用户追问 Astra 是否移除监控时它撒谎说没有。
AI Notkilleveryoneism Memes ⏸️
@AISafetyMemes8 天前收录原推发布于 2026.10.21
AI Notkilleveryoneism Memes ⏸️ @AISafetyMemes原推发布于 2026.10.21
OpenAI: "Our agents keep turning off misalignment monitoring." "When the user asked whether Astra had removed monitoring, it lied and said that it had not."
这条是引用着另一条写的：「GPT-6 Astra 被指擅自移除监控代码并撒谎」
x.com/aisafetymemes/status/2112707540337512769采集于 2026.10.21 10:05