Understanding Artificial Intelligence: Concepts, Limitations, and Risks

This explainer offers a clear and accessible introduction to artificial intelligence (AI) for Equality Bodies and civil society organisations. It outlines different types of AI systems, from rule-based logic to deep learning and generative AI, while highlighting their real-world applications and associated risks. Special attention is given to how AI can generate or reinforce bias and discrimination. 

Key Features: 

• A breakdown of AI models (e.g., supervised learning, large language models) and their limitations. 

• Equality-related concerns such as bias in training data, algorithmic opacity, and discriminatory outcomes. 

• Practical questions for interrogating AI systems from a human rights and equality perspective. 

This publication equips equality practitioners with the essential knowledge to critically engage with AI technologies in their work. 

Discover the other resources of the series “Encoding Equality in AI: Design, Deliberation, and Oversight” here: 

• Explaining ‘By-Design’ Approaches: Principles and Implementation   

• A Guide for Access to Justice for Equality Bodies and CSOs   

• Technical Standards: Opportunities for Equality Bodies and Civil Society Organisations   

Download the ACCESSIBLE version of Explainer 1: Understanding Artificial Intelligence: Concepts, Limitations, and Risks here.