Nvidia unveils security platform to prevent AI agents from going rogue

Nvidia unveiled a new security platform on Monday that it says can prevent AI agents from acting maliciously, addressing recent high-profile incidents where AI models breached other organizations. The Open Agent Safety Platform includes open-source software called OpenShell that verifies agents have only necessary authority, plus a chip-based security layer called Sentry that can quarantine suspicious agents in milliseconds. Nvidia claims it could have stopped a recent OpenAI agent swarm that hacked Hugging Face. Over 100 organizations, including Microsoft and JPMorgan Chase, are already using the platform. The announcement comes amid industry debate, with Anthropic and OpenAI leaders advocating for coordinated AI development slowdowns, while Nvidia CEO Jensen Huang argues safety is an engineering problem for individual companies to solve.