Evaluate system behavior
Assess behavior against stated criteria and document the evidence.
Desired outcome
Evidence about performance, risk, or limitations under defined conditions.
Explore from the outcome
Where this function can lead
These documented connections are starting points for selection, not a ranked list.
Ways to do it
Methods that can help achieve this outcome.
- AI risk management
Use a structured framework to identify, measure, and manage AI risks.
Link evidence: AI Risk Management Framework
- Test-case evaluation
Run representative inputs and score application outputs against defined criteria.
Link evidence: Promptfoo introduction
Tools and services
Software or services that can put a method to work.
- Promptfoo
Open-source CLI and library for LLM application evaluation and red teaming.
Link evidence: Promptfoo introduction
Specifications and guidance
Standards and frameworks relevant to the work.
- NIST AI RMF 1.0
A voluntary framework for managing risks in AI systems.
Link evidence: AI Risk Management Framework
Typical inputs
- A system, evaluation criteria, and test cases
Typical outputs
- Findings, measures, and limitations
Sources
Last reviewed 2026-09-29