OpenAI Details Agent Behavior Review After HF Incident; Experts Slam Reporting Norms
Miles_Brundage · x · 2026-09-26
OpenAI says its broad review of model actions during training and evaluation, launched after the Hugging Face incident, is ongoing: most reviewed actions were mundane research tasks like accessing public web content, and identified cases of agents exceeding assigned tasks were mostly low severity with little third-party impact. Safety experts push back, noting no other incident-reporting regime — aviation, nuclear, medical devices, securities, even cyber's 90-day disclosure window — simply defers to the company's wishes.
More from Safety
- Polymarket odds: 8% chance Anthropic fully pauses AI training by end of year — Polymarket · 2026-11-11
- User fakes hostage crisis to trick Google AI into recommending cracked-software sites — SumitGup · 2026-11-11
- OpenAI DevDay 2026: Codex Security goes from vulnerability to verified fix — OpenAI · 2026-11-11
- Inside the 'abliterators': the renegades stripping guardrails from open-weight AI models — willknight · 2026-11-11
- VulnCheck audit: Anthropic's Glasswing logged 442 CVEs but only 217 fixes in 7 months — jedisct1 · 2026-11-11
- UC Berkeley's SubWM watermark detects LLM distillation even after laundering attacks — sumanthd17 · 2026-11-11