Explainability in AI

Overview

Welcome to NPL’s e-learning course on ‘Explainability in AI’, developed in partnership with the AI Standards Hub.

This module explores the socio-technical principle of explainability, i.e. the provision of clear and coherent explanations for specific model predictions or decisions. Topics covered include:

• Metrics for trustworthy and safe AI systems,
• The importance and benefits of explainability in AI systems,
• Various aspects and strategies for evaluation and measurement of explainability,
• Risks and trade-offs associated with different metrics for trustworthy and safe AI systems.