Tampa General Hospital adds Nemotron-based safety checks on AI requests

NVIDIA says Tampa General Hospital is adding a safety check that screens AI requests before they are sent.

The check uses NVIDIA Nemotron, which Palantir and NVIDIA are post-training to review LLM calls. NVIDIA says the model is being trained to catch prompt injection, potential patient data leaks, and requests that should not go through. The source is an NVIDIA AI post on X; no results or deployment status have been reported.