Secure AI Agents From Test to Deployment
On September 28, 2026, NVIDIA announced the NVIDIA Open Agent Safety Platform, described as an open software platform and reference system design intended to strengthen AI-agent security from testing through deployment. NVIDIA says the platform provides full-stack governance and control across software and hardware.
On September 28, 2026, NVIDIA announced the NVIDIA Open Agent Safety Platform, described as an open software platform and reference system design intended to strengthen AI-agent security from testing through deployment. NVIDIA says the platform provides full-stack governance and control across software and hardware.
The announcement is relevant to teams building or deploying AI agents because it addresses security and governance across the agent lifecycle. The platform is announced, but the source does not establish that it is generally available or in preview.
NVIDIA’s announced Open Agent Safety Platform highlights a key shift for AI systems: security and governance must span the full agent lifecycle, from testing to deployment. For teams building digital workers, game agents, web experiences, and robotics-adjacent tools, reliable autonomy starts with controls designed in—not added later. martinezaistudios.com