UK Data (Use and Access) Act receives royal assent

The UK’s Data (Use and Access) Act received royal assent, ending a months-long standoff between the Commons and the Lords over whether the bill — otherwise a data-protection and data-sharing reform, unrelated to AI in its original scope — should also require AI developers to disclose what copyrighted material they used to train their models.

Baroness Kidron led the push for that requirement, tabling a transparency amendment that would have obliged AI firms operating in the UK to tell copyright holders when and how their work had been used in training. The amendment passed the Lords repeatedly over several rounds of parliamentary “ping-pong,” each time rejected by the Commons, where the government argued it pre-empted a separate consultation on AI and copyright that had closed earlier in 2025. The Lords eventually conceded rather than delay the bill’s other, uncontested data-protection provisions indefinitely.

As passed, the Act contains no new AI-specific transparency or licensing obligations. Instead it commits the government to publish an economic impact assessment of the options from its earlier consultation, and a report examining technical measures, licensing mechanisms and enforcement options for copyright in AI training — both within nine months of royal assent — plus an interim progress statement within six months, which the government duly published in December 2025.

The outcome left the underlying dispute over creative-industry compensation for AI training data unresolved and pushed to further government reports rather than statute, marking the UK, unlike the EU with its AI Act, as still without binding AI-specific transparency rules for training data. It was one of several 2025 flashpoints, alongside ongoing UK litigation such as Getty Images’ case against Stability AI, in the broader argument over whether copyright law needed to be rewritten for generative AI.