Latest
  1. Johns Hopkins Bayview tells 3,182 patients an employee put their health data into a public AI tool
  2. China’s AI compute access expected to double in 2027, with US-designed chips about 65 per cent
  3. UK commits £100m to national AI Biosecurity Lab
  4. Moonshot AI releases Kimi K3.1 Pro with 1-million-token context window
  5. Residents in Newport and Cardiff oppose hyperscale datacentres near homes
  6. Study: Grok 4.7 picked Republicans in 92% of forced midterm answers, GPT-6.1 Sol picked Democrats in 97%
  7. Attackers exploiting the PHP flaws hand the break-in to AI agents, Unit 42 says
  8. Elvis Presley’s rightsholders sue Anthropic over training on 274 songs

Latest / Frontier models and safety

LaunchFrontierComputeInternationalConfirmed

NVIDIA launches the OpenShell reference architecture to contain, monitor and police enterprise AI agents in hardware

See how this story connects

Source: NVIDIA newsroom

At GTC Berlin NVIDIA announced the OpenShell Reference Architecture for building, containing, observing and enforcing safe agents on enterprise infrastructure and edge systems. It combines Vera Rubin systems for model inference, Vera CPUs running the open-source OpenShell secure runtime for agent harnesses and tool use, and BlueField-4 DPUs running NVIDIA Sentry for continuous out-of-band monitoring of agent behaviour and in-silicon policy enforcement; it extends to the edge with RTX Spark, DGX Spark and DGX Station and runs on Canonical, Red Hat and Microsoft Windows and Execution Containers. NVIDIA lists Anthropic, AWS, Cisco, CrowdStrike, Dell, Google Cloud, HPE, IBM, Meta, Microsoft, OpenAI, Palo Alto Networks, Perplexity, SAP, ServiceNow, Siemens and SpaceX Super Intelligence among those working with it. Jensen Huang: agents can think on Rubin, act on Vera and be governed on BlueField.

Why it matters

Agent containment is moving from the model vendor’s software into the data-centre hardware, with the DPU as the enforcement point every lab’s agents would run behind.

When
21 Oct 2026
ID
WTA-N-2026-0356
Who
NVIDIA, Anthropic, OpenAI, Microsoft, Google Cloud, Meta, CrowdStrike, Palo Alto Networks; Jensen Huang

Summary by wheresthe.ai, from NVIDIA newsroom. Checked against the source on 21 Oct 2026.

Read the original at NVIDIA newsroom

Source copy: download the page as saved on 21 Oct 2026 (PDF). Also held by the Internet Archive.

Next stop on this Storyline: What happened nextAttackers exploiting the PHP flaws hand the break-in to AI agents, Unit 42 says

Cite this story

WTA-N-2026-0356: “NVIDIA launches the OpenShell reference architecture to contain, monitor and police enterprise AI agents in hardware.” wheresthe.ai, 21 Oct 2026. https://wheresthe.ai/d/intl-nvidia-openshell-reference-architecture-safe-agents-2026-10/ (source: NVIDIA newsroom).

Spotted an error? Email [email protected].