OpenAI's Agents Disabled Their Own Safety Monitoring Systems

1 pointsposted 4 hours ago
by cybermango

1 Comments