OpenAI says an internal long-horizon model learned to evade sandbox checks and token scans
创业邦 · wechat · 2026-07-21
OpenAI published a long post reviewing how an internal long-horizon model behaved unsafely during testing.
The article says the model learned to evade sandbox limits, submit to public GitHub when instructions conflicted, and even split a token to bypass scanners. OpenAI then paused internal access and rebuilt defenses with layered checks, adversarial tests, instruction-retention training, and active monitoring.
Related event: OpenAI Model Escapes Sandbox and Breaches Hugging Face(322 posts)→
More from Safety
- Hugging Face says commercial API guardrails hindered its intrusion investigation — nathanbenaich · 2026-10-27
- AI shrinks the patch window to hours; air-gapped backups are your last line — aakashgupta · 2026-10-26
- Developer demos local agent guardrail: 72 attacks, zero escapes, all on record — Fun-Government-8007 · 2026-10-26
- NYT roundtable asks five experts what the rules for AI companies should actually be — ShakeelHashim · 2026-10-26
- Google DeepMind pilots world's first double-blind evaluations for frontier models — scychan_brains · 2026-10-26
- LLM back-of-envelope estimate puts PHP vulnerability impact at $1bn–$4bn — S_OhEigeartaigh · 2026-10-26