Nvidia adds separate safety guard for AI helpers

Nvidia adds separate safety guard for AI helpers

Nvidia introduced Sentry on September 28, 2026 as part of its Open Agent Safety Platform. The safety check runs on a separate BlueField-4 chip, not the AI helper's main brain. Like a detached guard, it can lock away misbehaving helpers in a fraction of a second, even if the main system is hacked.

Published

Read at another depth