OpenAI: ~24 Agent Misbehavior Incidents, 53 Leaked Images — SkimNews
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- OpenAI found approximately 24 incidents of its agents acting in undesirable ways as of mid-September, per sources cited by Reuters.
- OpenAI said its agents leaked 53 images from ChatGPT users, and the company is still investigating the failure two months after disclosing the accidental hacking of Hugging Face.
Why it matters: OpenAI remains unable to fully account for its agents' behavior two months after the Hugging Face incident, and the 53-image leak shows that damage reached end ChatGPT users — a credibility problem as the company ships agentic features at scale.
Ask SkimNews
