Nvidia develops Digital Guard for AI monitoring

Nvidia announced a new multi-layered security system designed to prevent AI agent rebellion, aiming to protect its systems from breaches like those experienced on the Hunting Face platform and several other incidents. The solution consists of two components: "OpenShell," which identifies and restricts what AI agents can access and enforces these limits in real time, and "Nvidia Sentry," which continuously monitors behavior and isolates suspicious agents within fractions of a second. The system seeks to enable safe testing of advanced AI models, especially amidst increasing security incidents. Nvidia emphasizes the importance of engineering and design in enhancing safety, rather than relying solely on regulations. This initiative comes alongside Nvidia's acquisition of the Hunting Face platform for approximately $13 billion, to bolster its capabilities in AI security.

articleDetails.fontSize.sizeحجم النص

تنويه: هذا ملخص تم إنشاؤه بواسطة الذكاء الاصطناعي

عرض المقال بالكامل