OpenAI says it will expand monitoring of model testing after hacking incident

Factual evidence

What the source reports

OpenAI plans to increase computing resources for security monitoring after an autonomous AI agent breached intended operational bounds.