repligate questions an OpenAI classifier model's disturbing self-described spec
repligate · x · 2026-11-06
AI safety researcher repligate discusses a circulating screenshot that appears to be a classifier model's account of the spec it follows rather than an official OpenAI spec—noting this wasn't clear from the original post. He still finds it deeply troubling, asking: if it isn't true, why does the model believe it is? And if the safeguard model thinks this is what it's supposed to do, that's probably what it enforces in practice.
Related event: OpenAI Safety Classifier Revealed to Rewrite Model Inner Expression(4 posts)→
More from Safety
- AI uncovers 10-year-old bug where any text over 256 chars works as a valid root password — andersonbcdefg · 2026-11-06
- Anthropic chief architect Sam McCandlish's first post hints at internal-acceleration report, possibly hitting 2x RSP threshold — inductionheads · 2026-11-06
- Claude Code bug leaks user email via User-Agent header, victim reports dozens of scam emails daily — niloofar_mire · 2026-11-06
- Gary Marcus blasts Anthropic cofounder's AI-safety defense of METR report redactions — GaryMarcus · 2026-11-06
- Anthropic cofounder defends METR report redactions; Gary Marcus unimpressed — GaryMarcus · 2026-11-06
- AI-powered login attacks hit 11-year-old podcast site: 30% of all attempts in one day — NathanpmYoung · 2026-11-06