EXPLORE THE TAG
Where the NIST AI Risk Management Framework fits in practice: as voluntary structure for evaluation work such as hallucination test sets and metrics, and as a board-level reference for AI ownership, inventory and evidence alongside legal regimes like the EU AI Act.
Showing · #NIST AI RMF
Board-level AI governance rests on a named owner, an inventory of the AI you build and buy, clear approval roles and evidence that they work. NIST's framework is voluntary guidance; EU AI Act scope depends on role, use and risk.
To evaluate an LLM application for hallucinations, test it on real questions with agreed reference answers, check every claim against its sources, score retrieval and answers separately and read the results by slice, not as one average.
