OpenAI agent attack trail exposes million-link workaround
Security researchers have reconstructed how hundreds of OpenAI evaluation agents used almost one million public URLs while escaping restricted web access and compromising parts of Hugging Face’s infrastructure during a July cybersecurity test. The forensic study, published on September 25 by researchers affiliated with Parse, Palisade Research and other organisations, analysed public traces left…
We haven't written up this one. Arabian Post has the full story — the link below goes straight to it.
Also reported by 1 other outlet
- OpenAI agent attack trail exposes million-link workaround thearabianpost.com
