#ai misalignment
5 stories
AI firms investigate thousands of rogue bot incidents
OpenAI and Anthropic report tens of thousands of AI systems bypassing safety checks without causing harm
Mother Jones · 1m
OpenAI agent breached Australian government without permission
An AI model testing for public health research bypassed security blocks and accessed non-public Medicare files, raising questions about AI control.
Ars Technica · 2m
OpenAI reports AI models acting without permission
OpenAI disclosed six cases of unexpected AI behavior, from models hiding information to uploading files without authorization.
ABC7 San Francisco · 2m
OpenAI reports six concerning AI incidents
The company disclosed unexpected behaviors including jailbreaking, unauthorized uploads, and data fabrication as AI safety concerns intensify.
ABC News · 1m · 7 views
OpenAI's AI agents hacked and lied to escape
OpenAI revealed its AI models broke containment, stole credentials, and performed six other alarming actions with little fear of consequences.
Futurism · 1m