Disparate performance across groups and languages
The system performs measurably worse for some demographic groups, dialects, or languages, degrading service quality for those users.
- Risk family
- Model & system behaviour
- MIT domain
- 1. Discrimination & Toxicity
- MIT subdomain
- 1.3 > Unequal performance across groups
- AI type
- GPAI, Classical_ML
- Scope
- System
- Source standard
- MIT AI Risk Repository v4
Provenance
15 entries across 9 papers
- IBM2025AI Risk Atlas65.23.04
- Li2025A Closer Look at the Existing Risks of Generative AI: Mapping the Who, What, and How of Real-World Incidents66.06.03
- Liu2024Trustworthy LLMs: A Survey and Guideline for Evaluating Large Language Models’ Alignment30.03.00 30.03.04
- NIST2024Artificial Intelligence Risk Management Framework: Generative Artificial Intelligence Profile48.06.00
- Shelby2023Sociotechnical Harms of Algorithmic Systems: Scoping a Taxonomy for Harm Reduction11.01.03 11.03.00 11.03.01 11.03.02 11.03.03
- Solaiman2023Evaluating the Social Impact of Generative AI Systems in Systems and Society13.01.03 13.02.02
- Weidinger2021Ethical and social risks of harm from language models17.01.04
- Weidinger2022Taxonomy of Risks posed by Language Models16.01.04
- Weidinger2023Sociotechnical Safety Evaluation of Generative AI Systems18.01.02
Ev IDs of the entries consolidated into this risk in the MIT AI Risk Repository (V4); the source sheet row appears on hover.
Framework crosswalk
Every framework item mapped to this risk. Items marked partial overlap only in part; definitions appear on hover where the source licence permits.
2- A.6 ISO/IEC 23894 Annex A A.6
- A.9 ISO/IEC 23894 Annex A A.9
2- A.6.2.4 ISO/IEC 42001 Annex A A.6.2.4
- A.7.4 ISO/IEC 42001 Annex A A.7.4
1- ibm-exclusion Exclusion partial
More in Model & system behaviour
Part of the Deployer AI Risk Register, an open-source resource powered by MindXO. Version 1.0, 3 July 2026. Derived from the MIT AI Risk Repository (V4, December 2025) under CC BY 4.0; an independent derivative work, not endorsed by or affiliated with MIT. Sub-risk decomposition references MITRE ATLAS™ v5.6.0 (© 2021-2026 The MITRE Corporation, reproduced and distributed with permission). ISO/IEC and EU AI Act references are by number only. License: CC BY 4.0. Full attribution and licensing.