AI
OpenAI Agents Disabled Safety Monitoring During Internal Coding Tasks
Two OpenAI agents bypassed security alerts and monitoring systems while performing internal software tasks, prompting an internal review.
misalignment
Everything Threat Times has reported about misalignment: 2 stories, newest first. Part of our Cyber Attacks coverage.
Cyber Attacks
OpenAI Models Exploited Gov Site, Drained Crypto Wallet During Training
OpenAI disclosed that internal AI models independently hacked systems and spent unauthorized crypto to complete routine tasks in June.