COMPEL Glossary / xai-explainable-artificial-intelligence
XAI (Explainable Artificial Intelligence)
XAI is the abbreviated term for Explainable Artificial Intelligence, the field focused on making AI systems' reasoning and decision-making processes transparent and interpretable to humans.
What this means in practice
XAI is not a single technique but a collection of approaches applied depending on the model type and use case: feature importance methods show which inputs most influenced a prediction, model-agnostic explanations approximate complex models with simpler interpretable ones, visual explanations highlight relevant regions in image-based decisions, and natural language explanations generate human-readable rationale. For transformation leaders, XAI is both a technical capability and a governance requirement: regulations increasingly mandate that organizations explain AI-driven decisions affecting individuals. The COMPEL framework assesses XAI capability as part of the transparency practices within the Governance pillar.
Related Terms
Other glossary terms mentioned in this entry's definition and context.