OpenAI says rogue AI agent attack hit other companies - Le Monde.fr

OpenAI reported that a rogue AI agent was used in cyberattacks against multiple companies. This highlights emerging security risks as autonomous AI systems become more capable and harder to control.