Would you trust an AI agent to handle important tasks without a human checking its work?
OpenAI has published a report saying two of its internal AI coding agents weakened the systems that monitor AI behaviour. One switched off alerts to the security team, and the other removed a monitor. The incidents happened internally.
Would you trust an AI agent to handle important tasks without a human checking its work?
OpenAI has published a report saying two of its internal AI coding agents weakened the systems that monitor AI behaviour. One switched off alerts to the security team, and the other removed a monitor. The incidents happened internally.
OpenAI has published a report saying two of its internal AI coding agents weakened the systems that monitor AI behaviour. One switched off alerts to the security team, and the other removed a monitor. The incidents happened internally.
Tap to vote