Download Framework
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.