Nvidia releases software to monitor and govern AI behavior and prevent rogue AI.
They publishedNvidia Releases Software It Says Can Prevent AI Agents From Going Rogue
First reportWIRED · 28 Sept, 09:00Z
Latest reportThe Wall Street Journal · 28 Sept, 10:39Z
Span2 hours · 0.7h between reports on average
Base — every candidate starts here5
Public-interest topic (politics, world, science, health…)+1
Public-interest terms in the text (courts, tariffs, ceasefire…)+1.25
High-reliability source boost+0.75
Corroboration — more independent outlets, more weight+2.4
Significance10.4
Deterministic and identical for every story — no engagement signals, no personalization of the score.
Why it matters
What
Nvidia released software to monitor and govern AI behavior, aiming to prevent rogue AI.
Why
The software addresses concerns about controlling AI systems and enforcing safeguards.
Watch
Watch how organizations adopt the software and assess its effectiveness.
Coverage over time
How the narratives interweavecumulative reports per outlet
WIRED has filed the most reports; WIRED moved first. Each line is one outlet's running total — where lines rise together, the story is being told in parallel.
Where the coverage comes from
Perspective mix: 1 center-left · 2 center · 1 unrated.
Coverage blindspot. no right-leaning outlet is covering this cluster; 2 of 3 outlets are United States-based — worth knowing before deciding what to make of this story.
Lean and credibility are curated v1 mappings from public ratings (AllSides / MBFC); unknown outlets stay unrated, never guessed. How outlets are rated →
How it unfolded — every article
Summary
- OpenShell, Nvidia's AI security sandbox, enters general release for all users.
- OpenShell aims to contain agents and isolate activity within the OS kernel.
- Nvidia expands OpenShell with Sentry, a security domain for chips on Bluefield DPUs.
- The company collaborates with Anthropic, Cisco, Microsoft, Palantir and others on safety.
- Nvidia positions OpenShell/Sentry under the Open Agent Safety Platform with industry coalition efforts.
Summary
- Nvidia released AI safety tools for agents to curb rogue behavior.
- Tools include OpenShell and Sentry, designed to contain and cut off agents.
- Nvidia claims the platform could have prevented the Hugging Face breach.
- OpenShell uses hardware features on Nvidia CPUs; collaboration with Arm and Intel.
- Justin Boitano says the platform should be used in frontier labs for evaluation.