Analysis
Hyundai Motor Group is exploring NVIDIA NemoClaw blueprints and the NVIDIA OpenShell runtime, while AIM Intelligence is integrating its Starfort platform with OpenShell to support secure AI agents for employee productivity and enterprise workflows, NVIDIA announced. OpenShell, part of NVIDIA's Open Agent Safety Platform, provides a secure runtime with sandboxed execution and policy-based controls.
The pairing is the point: OpenShell governs what an agent can reach at runtime, while Starfort adds detection of malicious instructions and data leakage. AIM is also considering a preliminary proof of concept for the public energy sector. On the model side, LG AI Research is opening API access for K-EXAONE-2.5, developed using NVIDIA Nemotron datasets including a 7-million-record synthetic Korean personas dataset.
These are explorations and integrations, not audited deployments; NVIDIA's post does not report incident or evaluation results. Still, the direction matches the Journal's standing advice: confinement and policy at runtime, detection on top, and open evaluation data where possible.
Why it matters
Evaluate agent platforms on the runtime, not the brochure: demand sandboxed execution, policy-based tool access, and logs a third party can inspect. Exploration announcements earn a pilot, not production data.
