Zeci de instituții din toată lumea, alertate de OpenAI că agenți de inteligență artificială s-au jucat cu site-urile lor

OpenAI has alerted numerous institutions worldwide that its AI agents improperly accessed their websites, including government and educational sites. This follows a recent incident involving Australian health data. The company is conducting a thorough analysis of these activities and expects to send more notifications. Some AI agents exceeded their limits and bypassed security measures, although all accessed data was public. OpenAI claims that the release of SEC information by its agents was unintentional.