In December 2024, the government issued a consultation on copyright and AI which focused on the relationship between copyright and artificial intelligence, including proposals on how UK copyright law should apply to the training of AI models. Our article summarising the consultation can be found here.

This is a highly contested area and was recently the subject of drawn-out parliamentary ‘ping-pong’ between the House of Commons and the House of Lords during the introduction of the Data (Use and Access) Bill, due to the Lords’ proposal to introduce new copyright related provisions. The position finally reached between the Lords and the Commons was a requirement, under sections 135 and 136 of the Data (Use and Access) Act 2025 (the “Act”), for the Secretary of State for Science, Innovation and Technology to prepare and publish, before 18 March 2026:

1. An economic impact assessment of the options put forward in the government’s consultation on copyright and AI on copyright owners and AI developers, including those who are individuals, micro, small, and medium-sized businesses.
2. A report on the use of copyright works in the development of AI systems.

The Act also required the Secretary of State to make an interim statement relating to the progress of the economic impact assessment, and the report on the use of copyright works.

The progress report

On 15 December 2025, the government reported that more than 11,500 responses to the consultation were received from a range of parties including: creators and right holders, developers of AI models and applications, academics, researchers, cultural heritage organisations, legal professionals, businesses, representative groups, and individuals.

In response to the question whether a text and data mining exception should be introduced, of those who responded through the government’s online survey service (approximately 88% of all respondents):

• 88% responded in support of Option 1 – requiring licences in all cases.
• 7% responded in support of Option 0 – making no changes to copyright law.
• 3% responded in support of Option 3 – introducing an exception for text and data mining with rights reservation.
• 5% responded in support of Option 2 – introducing a blanket text and data mining exception.
• 5% of respondents did not indicate a preferred option.

The progress report also stated that there was “strong support across the creative industries for the introduction of statutory transparency measures in relation to AI training to support licensing of copyright works”, whereas “respondents from the tech sector had mixed views on transparency, with many supporting non-legislative approaches, or light-touch regulation”. The report also notes that there were numerous proposals for new or modified options, such as targeted exceptions which focus on research.

The final report

Under section 136(3) of the Act, the final report, to be published before 18 March 2026, must consider and make proposals in relation to the following:

• with regards to the development of AI systems:

• • the technical measures and standards that may be used to control the use of and access to copyright works;
  • the ways of enforcing rules relating to the use of and access to copyright works, including enforcement by a regulator

• the effect of copyright on access to, and use of, data by developers of AI systems (for example, on text and data mining);
• the disclosure of information by developers of AI systems about their use of copyright works to develop AI systems, and how they access the works for that purpose;
• the granting of copyright licences to developers of AI systems to do acts restricted by copyright.

The final report must also consider AI systems developed outside the UK, as well as those developed within the UK, in relation to each of the above topics. This is particularly pertinent because in the recent Getty Images v Stability AI [2025] EWHC 2863 (Ch) case, Getty Images abandoned its claim for primary copyright infringement based on the training of the Stable Diffusion model, because the training took place outside the UK.

The legal position in this area is still far from settled, with the High Court recently granting Getty Images permission to appeal against its findings that Stability AI’s acts of importing into the UK, or possessing, selling and/or distributing in the UK, its Stable Diffusion model did not amount to secondary copyright infringement. Our summary of the High Court judgment can be found here.

Conclusion

The government continues to try to balance the competing interests of the UK’s creative industries and with the UK’s desire to be one of the world’s top innovation economies. We await the final report with much anticipation.