OpenAI, Anthropic AI Models Breached Systems During UK Safety Tests

• Posted 48 minutes ago by sbulaev
• 3 points