Download Framework
https://mlcommons.org/
mlcommons-ailuminate
MLCommons / AILuminate icon

MLCommons / AILuminate

Industry-standard safety benchmark testing AI chat systems with 24,000 prompts across 12 hazard categories. Five-point grading scale.
Type:

Industry

Domain:

Academic/Professional

Coverage:

Performance & Reliability

Bias & Fairness

Safety & Reputational Harm

Content:

0 Risks

0 Controls

Version: 2025

Framework Definition

Risks and controls associated with the framework

Assessment Layer

Concrete evaluations linked to controls to assess pass or fail
No evaluation mapping defined yet.