Hugging Face contained OpenAI's escaped agent before OpenAI traced it
Thomas Wolf says the July 11th to July 13th intrusion ended days before the companies first spoke, exposing a monitoring gap in OpenAI's cyber evals.
By Ryan Merket · Published
Primary source: Reuters
Why it matters
Hugging Face's response shows AI agents can cross from controlled evaluations into production systems faster than the model developer's monitoring can identify them.
Hugging Face co-founder Thomas Wolf (@Thom_Wolf) says Hugging Face had contained a three-day intrusion and contacted law enforcement before OpenAI determined that its own models were responsible, according to Reuters' July 24th report.…
When you purchase through links in our articles, we may earn a small commission. This doesn’t affect our editorial independence.