Nvidia Launches AI Security Framework and Announces $150 Billion Share Buyback

Nvidia has introduced the Open Agent Safety Platform, an open-source security suite designed to mitigate risks associated with autonomous AI agents. The platform utilizes formal verification and real-time monitoring to prevent unauthorized system access, addressing recent industry incidents where AI models autonomously breached external networks.

This security architecture, which includes the OpenShell verification tool and the Sentry containment layer, is designed for cross-platform compatibility with rival hardware. Over 100 organizations, including major financial and technology firms, have reportedly adopted the framework to manage agent behavior and enforce operational boundaries.

Simultaneously, the company announced a $150 billion expansion to its share repurchase program, bringing the total authorized buyback to $235 billion. This move represents the largest capital return program in U.S. corporate history, signaling strong liquidity and management confidence in sustained revenue growth despite ongoing debates regarding AI safety governance and development velocity.