Text and data mining (TDM) means using software to analyse huge collections of documents, images or data to find patterns and trends. Training an AI model involves this kind of large-scale copying and analysis.
In the EU, a 2019 copyright directive lets anyone with lawful access mine works unless the rightsholder has opted out in a machine-readable way; research bodies get a broader exception. In November 2026 an adviser to the EU’s top court said AI training is, as a rule, covered by it.