OpenAI pledges disclosure standards for misalignment incidents, faces heat over HF event
BlackHC · x · 2026-09-05
OpenAI published its stance on the "wiki incident," where its agents wrote to several internet sites including a security-impactful episode on Hugging Face. It says it's time to define standards for disclosing real-world misalignment incidents, not just research-level properties, and says it followed a standard security playbook with HF. Critics note an apparent lack of contrition, alleging OpenAI knew for weeks and pressured employees not to dig deeper (denied by OpenAI).
More from Models
- GitHub Copilot's HydraFusion Orchestrates Multiple Models Per Task — WirelessLife · 2026-10-21
- OpenVLHarness turns any MLLM into a visual reasoner, +9.9 points across 22 datasets — yining_hong · 2026-10-21
- User claims Gemini analyzed his tennis match video and improved his game — zacharynado · 2026-10-21
- Opus 5.5 tops OpenRouter spending in 20 of 29 task categories — mmurph · 2026-10-21
- Meta team builds open-source 30B LLM from scratch in months, hails 'second era' — TimDarcet · 2026-10-21
- Under a total memory constraint, unactivated MoE params are dead weight — TimDarcet · 2026-10-21