The CFAA says knowingly. Negligence is by definition an excuse for that.

gmerc 14 hours ago | parent | next [-]

Removing the guardrails is knowingly given the risk

kibibu 7 hours ago | parent | prev [-]

I think there's a reasonable case that the agent knew it was breaking into the system, for some definition of knew.

I think OpenAI would be very reluctant to let this go to a place where the reasoning was part of discovery.

	▲	wbl 6 hours ago \| parent [-]
		Agents aren't subjects of criminal law. I agree there may be civil liability, I know far less about that.