UK Copyright and AI Text and Data Mining

Text and data mining (TDM) uses automated analysis to find patterns in text or data, and often requires copies. Section 29A of the Copyright, Designs and Patents Act 1988 permits copies for computational analysis where the sole purpose is non-commercial research and the researcher already has lawful access. Sufficient acknowledgement is required unless impractical. Contract terms preventing qualifying copies are unenforceable, while providers may take reasonable measures to protect network security or stability. This narrow exception is not a blanket route for commercial model training. If a TDM purpose is commercial, section 29A does not itself authorize the copying; permission or another applicable legal basis may be needed. It is also too broad to claim every commercial training use necessarily infringes: the work, acts, permissions, other exceptions and facts matter, and cross-border activity raises territorial questions. The government’s report of 18 March 2026 reviewed policy options from its 2024 consultation. It said opposition, evidence gaps and a changing market meant the broad copyright exception with opt-out was no longer its preferred way forward. Instead, government proposed gathering further evidence and considering alternative interventions. This report states policy direction; it did not enact a new TDM exception. Current analysis therefore starts with existing copyright law, permissions and case-specific legal questions while courts and parties resolve disputes. Organisations should avoid substituting policy announcements for enacted rules.