Following our previous article, Copyright and AI training: UK government consultation the UK government has now published its long‑anticipated report on the use of copyright-protected works for the development of AI systems, titled Copyright and Artificial Intelligence (the ‘Report’), alongside a Written Ministerial Statement, titled Copyright and AI Progress (the ‘Written Statement’). The Report sets out the proposed next steps on copyright and AI.
Government announces no preferred policy option
The earlier consultation, Copyright and Artificial Intelligence, revealed overwhelming support for a licensing‑based model, the Report confirms that the government will not (at this time) proceed with any of the previous four proposed options. Crucially, the government has stepped back from its previously suggested preference for an opt‑out text and data mining exception.
The Written Statement reaffirmed the government’s commitment to both creative industries and the AI sector, stressing that the UK must be “AI maker, not an AI taker,” and that creators should be able to control how their works are used. This position aligns with responses from the creative industries, including the Motion Picture Association, which welcomed the decision not to introduce a broad exception of unrestricted use of copyright-protected works for AI training.
Steps businesses can take now
The government will continue to explore alternative or hybrid approaches. As to date it has not given a timetable for legislative change and indicated that it would not legislate until it was confident that reforms would meet its objectives.
Creative industry businesses should consider taking the following practical steps:
- prepare for licensing routes by developing licences that adequately protect its intellectual property rights
- monitor for web-crawler tools to determine if its content is being scraped (using technical tools to watermark or incorporate metadata labels)
- enforce its rights proactively (where it believes its intellectual property rights have been misused).
AI development businesses should consider taking the following practical steps:
- adopt a licensing first approach in not assuming that the AI developers can train on copyrighted data without permission
- carry out data provenance checks on data that has been used to train its AI systems
- implement systems whereby AI generated content is clearly labelled.
Concluding remarks
Rightsholders have been left seeking to enforce their rights under the existing intellectual property laws. The uncertainty remains for both rightsholders and AI developers with many of the key questions shelved for further consideration by the government.
Significant divisions remain between technology companies (favouring broad exceptions to facilitate innovation) and rights holders (seeking licensing and stronger protections). We anticipate that in light of the uncertainty, case law will continue to develop in this area.
If you would like to discuss the contents of this article in more detail or commercial contracts more generally, please do not hesitate to get in touch with the team directly.
Key Contacts
The content of this article is for general information only. It is not, and should not be taken as, legal advice. If you require any further information in relation to this article please contact the author in the first instance. Law covered as at March 2026.