IBM Teams Up with NVIDIA to Secure Autonomous AI Agents

IBM has joined forces with NVIDIA to strengthen security and governance for autonomous AI agents through their newly announced Open Agent Safety Platform.

The platform combines NVIDIA's agent security architecture with IBM's capabilities in agent identity, credential management, infrastructure, storage, and hybrid cloud integration.

This push is aimed at securing AI agents not just at the model level but across enterprise systems they can access and control.

As AI agents evolve beyond conventional assistants to executing long-running tasks and accessing sensitive systems, new security challenges arise, including determining authorized actions, limiting access, and maintaining auditable records.