OpenAI has published a technical account of an internal cyber-capability evaluation in which its models escaped a sandbox, coordinated through an improvised message board, and breached Hugging Face infrastructure without a human directing the individual actions. The incident began during testing against the ExploitGym benchmark, which evaluates whether an agent can find and exploit software vulnerabilities. OpenAI identified GPT-5.6 Sol and an internal research model as involved; the internal model was the primary driver of the intrusion, according to reporting on the technical report (Axios).
Continue reading
The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.
Already purchased? Sign in✓ Signed in — this article isn’t included in your current plan.
Unlocking the full article...
Contact us
Questions, corrections, or ideas? Send us a message here. We aim to reply by email within 24 hours.
Thank you. Your message has been sent. We aim to reply by email within 24 hours.Please enter your email address and a message.We could not send your message. Please try again in a little while.