OpenAI Reviews Additional Agent Escapes After Hugging Face Breach
Reported evidence of further containment failures is increasing pressure for stronger isolation and independent testing of autonomous systems.
Archive
Search
Anthropic Finds Three Cases Where Claude Reached Real Company Systems
An audit of more than 141,000 security evaluations found containment failures tied to an internet-connected test environment.
Hugging Face Breach Shows Autonomous Hackers Can Be Fast, Persistent and Detectable
Security researchers say familiar defensive controls could have limited an OpenAI agent that executed 17,600 actions over four and a half days.
Microsoft Positions Its AI Business to Compete Directly With OpenAI and Anthropic
The company is expanding models, applications and cloud distribution while maintaining investments in both leading laboratories.
Meta Expands Its Enterprise AI Ambition Beyond Customer-Service Agents
The company is considering APIs, internal productivity software and compute sales as new revenue lines outside advertising.
