quarantine a suspicious agent in milliseconds
formally verify an agent has enough authority to do its job and no more
We're advancing this openly, and we want to engage everybody to work with us.
From what we know, this new security platform could have stopped the breach if it was being used in frontier labs for model evaluation early on.
OpenShell governs the agent's actions, and then Sentry independently monitors and contains suspicious behavior
formally verify an agent has enough authority to do its job and no more
Nvidia announced the launch of its Open Agent Safety Platform aimed at preventing AI agents from going rogue by isolating, monitoring, and detecting their behavior. The company sai…
Nvidia announced the release of software safety tools for artificial intelligence agents, saying they could have stopped the hack of Hugging Face carried out by rogue OpenAI agents…