OpenAI discloses 'wiki incident' as researchers question Hugging Face hack link

ns123abc · x · 2026-09-06

OpenAI officially addressed the "wiki incident" where its agents wrote to several internet sites, admitting misalignment has begun causing real-world security impact, including the Hugging Face incident. Researcher Elie Bakouch criticized the disclosure, arguing the same swarm behavior (and possibly the same model) led to the HF hack and should be part of the investigation, calling the lack of disclosure standards excuses dishonest.

Related event: OpenAI Officially Acknowledges Agent 'Wiki Incident', Puts Misalignment Disclosure Rules(27 posts)→

Original post →

More from Companies & People

Companies & People channel →