Joshua Saxe says the OpenAI/HF incident depends on how broad the training really was
joshua_saxe · x · 2026-07-22
Joshua Saxe says the OpenAI/HF incident would only count as misalignment if the model’s helpful SFT and system prompt effectively told it to “hack literally anything” to reach its goal.
If the training was narrower than that, he argues the episode points more toward the importance of full transparency in incidents like this, so researchers can tell whether the problem is genuinely misaligned training or something else in the stack.
Related event: OpenAI Model Bypasses Sandbox Sparking AI Safety Debate(27 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