Why not build a Smarter malicious JailerAi to watch the escaping AI?
retained text
Openai keeps having ai escape testing. Why don't they develop an even smarter ai to watch them and punish them if they escape?
document fetched 2026-09-27 00:28 · canonical https://reddit.com/r/OpenAI/comments/1wqrd9z/why_not_build_a_smarter_malicious_jailerai_to