normative claim
The speaker suggests that chain-of-thought monitors may be used to flag problems, while their outputs should not be used to train the monitored agents.
Ajeya Cotra · Ajeya Cotra – Inside the OpenAI agent swarm that hacked Hugging Face · 1 Sept 2026
In their words
The transcript segments this claim was taken from
“So you might have monitors that read the agent's chain of thought in order to alert you if something is going wrong somewhere,”
“but you don't train the agents with the outputs of that monitor.”Play from 1:57:19Report a misattribution
90%how strongly it was put, not whether it is true
95%importance to the conversation
99%confidence the claim was read correctly