Skip to main content

Zero Trust

NVIDIA Launches Open Agent Safety Platform With CrowdStrike

NVIDIA Launches Open Agent Safety Platform With CrowdStrike

Austin – October 12, 2026 -- NVIDIA introduced the NVIDIA Open Agent Safety Platform, an open source platform and reference design that applies multi-layered security controls across the software, runtime and infrastructure where AI agents operate, with CrowdStrike collaborating on the effort.

NVIDIA Sentry enforces policy from BlueField-4 DPUs and can quarantine agents in milliseconds

At the software layer, NVIDIA OpenShell provides an open source secure runtime that governs how agents execute, what they can access and where inference occurs. NVIDIA Sentry adds an independent watchdog running on NVIDIA BlueField-4 DPUs. It uses NVIDIA DOCA for out-of-band, in-silicon monitoring of agent activity and policy enforcement.

Shared Signals Framework Brings Near Real-Time Access to Zero Trust

Shared Signals Framework Brings Near Real-Time Access to Zero Trust

Global – October 12, 2026 -- Re-validating an access token at every request would multiply the load on an identity provider (IdP) by roughly 1,000x, assuming a 24-hour token is used 1,000 times. That figure is why the OpenID Foundation's Shared Signals Framework (SSF) is gaining traction as the mechanism for keeping Zero Trust access decisions current.

Short-lived tokens cut IdP load to about 24x but degrade user experience

Shortening a token's life from 24 hours to one hour raises IdP burden by only about 24x. Users, however, face repeated redirects to the IdP and back, which degrades the experience. Neither approach lets an organization instantly revoke access after a termination or rogue agent activity, or react when device or task conditions change while a token is still valid.