OpenAI and Anthropic Reportedly Uncover AI Safety Incidents Beyond Prior Disclosures
OpenAI has paused a training run while it investigates reports of AI model misbehavior, according to two published accounts. OpenAI and Anthropic are reported to have…