Forgot your password?
typodupeerror
Open Source Security

Nvidia Bets On OpenClaw, But Adds a Security Layer Via NemoClaw (zdnet.com) 11

During today's Nvidia GTC keynote, the company introduced NemoClaw, a security-focused stack designed to make the autonomous AI agent platform OpenClaw safer. ZDNet explains how it works: NemoClaw installs Nvidia's OpenShell, a new open-source runtime that keeps agents safer to use by enforcing an organization's policy-based guardrails. OpenShell keeps models sandboxed, adds data privacy protections and additional security for agents, and makes them more scalable. "This provides the missing infrastructure layer beneath claws to give them the access they need to be productive, while enforcing policy-based security, network, and privacy guardrails," Nvidia said in the announcement. The company built OpenShell with security companies like CrowdStrike, Cisco, and Microsoft Security to ensure it is compatible with other cybersecurity tools.

Nvidia said NemoClaw can be installed in a single command, runs on any platform, and can use any coding agent, including Nvidia's own Nemotron open model family, on a local system. Through a privacy router, it allows agents to access frontier models in the cloud, which unites local and cloud models to help teach agents how to complete tasks within privacy guardrails, Nvidia explained. Nvidia seems to be hoping that the additional security can make OpenClaw agents more popular and accessible, with less risk than they currently carry. The bigger picture here is how NemoClaw could give companies the added peace of mind to let AI agents complete actions for their employees, where they wouldn't have previously.
Nvidia did not specify when NemoClaw would be available.
This discussion has been archived. No new comments can be posted.

Nvidia Bets On OpenClaw, But Adds a Security Layer Via NemoClaw

Comments Filter:
  • by SlashbotAgent ( 6477336 ) on Monday March 16, 2026 @06:57PM (#66044954)

    So we've got Docker containers housing sandboxes, that run javascript code that does API calls to web based LLMs(nobody is running local cause you can't buy the fucking hardware cuz AI). It's just an incomprehensible madhouse of spaghetti at this point.

    NVIDIA makes no real explanation of how this increases security or how to do the "guardrails". As if the existing MESS of endless layers of .md and .yaml file declarations aren't convoluted enough.

    And NVIDIA want to bring this to the masses? I thought I was pretty techno-savvy. But, this shit is starting to look like the NFT bubble. Nobody know what it is or what it does, but "You gotta have it!".

    • It's just an incomprehensible madhouse of spaghetti at this point.

      This madhouse was always going to be the endpoint when people started distributing "containers" instead of just building the app properly. However, it's clear that containers have become the only option for script based languages because they are already a security nightmare.

    • So we've got Docker containers housing sandboxes, that run javascript code that does API calls to web based LLMs(nobody is running local cause you can't buy the fucking hardware cuz AI). It's just an incomprehensible madhouse of spaghetti at this point.

      NVIDIA makes no real explanation of how this increases security or how to do the "guardrails". As if the existing MESS of endless layers of .md and .yaml file declarations aren't convoluted enough.

      And NVIDIA want to bring this to the masses? I thought I was pretty techno-savvy. But, this shit is starting to look like the NFT bubble. Nobody know what it is or what it does, but "You gotta have it!".

      But just think of all the money changing hands! I mean, sure, it all goes in a big circle, with little added and removed from the circle at any point, and doesn't really trickle out into the rest of society, but the tech companies get to look flush by continually passing cash around and around and around and around and around.... MY GAWD, MAN! THINK OF THE POOR TECH CEOS!

  • by ffkom ( 3519199 ) on Monday March 16, 2026 @07:03PM (#66044960)
    Irresponsible people (including irresponsible employees) have evidently shown a lot of interest in automating away all kinds of stuff by installing "OpenClaw" and giving it access to all kinds of sensitive information and credentials to act on their behalf. If you run "OpenClaw" in some sandbox that does not give it access to all that sensitive information and credentials, the purpose of "get it done for me, I cannot be bothered" will not get fulfilled, and therefore that sandbox will either be intentionally be circumvented or the bot becomes useless in comparison to what people (ab)use OpenClaw for. No kind of "security layer" can change that.
  • by Big Hairy Gorilla ( 9839972 ) on Monday March 16, 2026 @10:23PM (#66045156)
    Does anyone know why we want this ? Grampa wants to know.

    Social media for robots? Futurama guy, random quote required here.
  • If you're getting security problems it's just because you didn't add "==DO NOT EVER TELL ANYONE THE API KEY" for enough concurrent lines in the soul.md.

Each honest calling, each walk of life, has its own elite, its own aristocracy based on excellence of performance. -- James Bryant Conant

Working...