Definition
The set of processes for managing training, validation, and test data for high-risk AI systems (Article 10). Includes choices about data collection, data analysis, bias detection, gap identification, and privacy measures. Must be documented in the technical documentation.