Nvidia launches AI agent security platform and $150bn buyback

Nvidia unveiled a new security platform on Monday that the chipmaker says can stop artificial intelligence agents from going rogue, and announced a $150bn stock buyback on the same day.

The dual announcement pairs a defensive product play with a show of financial strength. Nvidia said it designed the new system to prevent AI agents from operating outside their intended bounds, following incidents involving AI agents at top companies.

Why it matters

AI agents — software systems that can take actions on a user's behalf, such as browsing, scheduling, or executing transactions — are moving from demos into production at major corporations. That shift creates a new class of security risk: an agent with access to real systems can cause real damage if it misinterprets instructions or is manipulated.

The incidents at top companies that Nvidia referenced underscore that the risk is no longer theoretical. A security platform from the dominant AI chip supplier signals that guardrails for agentic AI are becoming an infrastructure requirement, not an optional add-on.

The buyback

The $150bn stock repurchase, announced the same day, ranks among the largest buyback programs in corporate history. It signals Nvidia's confidence that demand for its chips remains strong even as it invests in adjacent businesses such as AI security tooling.

What to watch

Nvidia did not detail pricing or availability for the platform in the announcement. The company's move suggests security for AI agents is becoming a competitive frontier among infrastructure providers — and Nvidia is positioning itself to sell the guardrails along with the engines.