GENAI-MEASURE-2.1: Evaluate generative AI bias and fairness across protected categories
Vallum evidence mapping for NIST GenAI Profile GENAI-MEASURE-2.1 — evaluate generative ai bias and fairness across protected categories. Documents evidence produced, gaps, and customer responsibilities.
Source Attribution
- Framework
- NIST GenAI Profile 1.0
- Control ID
- GENAI-MEASURE-2.1
- Source Organization
- NIST
- Official Source
- Generative AI Profile ↗
- Control Objective
- Evaluate generative AI bias and fairness across protected categories
Evidence Vallum Produces
Partial evidence through operation receipts and evidence-package metadata relevant to this control's objectives
Evidence Vallum Does NOT Produce
Training data provenance records, content filtering effectiveness metrics, or bias evaluation results
Known Limitations
Vallum provides evidence relevant to this NIST GenAI Profile requirement but does not constitute complete control implementation. Full compliance requires complementary controls at infrastructure, organizational, and procedural levels. Evidence covers Vallum-managed operations only and does not extend to customer applications, third-party systems, or physical infrastructure.
Your Responsibilities
Document training data sources, implement content safety measures, conduct bias evaluations, and maintain human oversight procedures
External Controls Required
Identity provider, infrastructure security controls, organizational policies and procedures, third-party assessments, and legal/regulatory counsel
Residual Risk
Evidence gaps exist between Vallum-managed operations and full control implementation. Time-of-check to time-of-use windows, infrastructure-level vulnerabilities, and organizational process failures remain outside Vallum's evidence boundary.
Relevant Titan Components
Need a detailed evidence mapping for your specific compliance program?
Request a Control Evidence Mapping