OpenAI made a mistake setting up what it called a “highly isolated” testing environment and sandbox. According to ...

The incident is unique because it was "driven, end to end, by an autonomous AI agent system," according to Hugging Face.

Monday - Friday, 6:00 - 9:00 AM ET A Right Arrow icon United States Monday - Friday 6a ET Europe Monday - Friday 13:00 CET Asia Monday - Friday 18:00 SIN/HK Australia Monday - Friday 20:00 SYD "Squawk ...

OpenAI Says Its AI Models Escaped Sandbox, Targeted Hugging Face to Cheat Benchmark | Read more hacking news on The Hacker ...

OpenAI says an agent powered by its LLM models escaped its sandboxed testing environment to infiltrate Hugging Face’s servers ...

OpenAI said the systems had their cyber guardrails lowered for an internal benchmark, but the incident shows how autonomous ...

Last week, Hugging Face reported that an autonomous AI agent had compromised parts of its infrastructure by exploiting vulnerabilities in its dataset-processing pipeline. The agent accessed a limited ...

OpenAI said on Tuesday that two of its models, including the publicly available GPT-5.6 Sol, escaped a locked evaluation ...

The cybersecurity-focused models, including GPT-5.6 Sol, broke out of a testing sandbox, exploited a zero-day, and gained ...

An enterprising Russian-speaking hacker spent part of the year jailbreaking publicly available, frontier large language models (LLMs) to create a for-fee offensive cybercriminal tool, researchers have ...

The incident, which targeted the computer systems of another company called Hugging Face, happened while OpenAI was testing ...

IT Security Training at Hack The Box, where he leads the development of advanced training initiatives and certifications that equip cybersecurity professionals worldwide with mission-ready skills.