In this article we take a look at what to expect in 2026 in the area of AI and UK copyright, including key judgments, legislative proposals and other developments on the horizon.
Copyright consultation outcome
The question of whether use of copyright works to train AI models infringes copyright is the most important issue for copyright creators, users and practitioners at the moment. Last year, the UK government ran a consultation seeking input on a number of options that attempt to strike a balance between AI developers being able to use copyright works when training models, and the rights of copyright holders.
The government’s favoured option was stated in the consultation to be a data mining copyright exception, with a rightsholder opt-out. However, the creative industries were strongly against this option (see our earlier article on the government’s Statement of Progress, published in December). And the objections appear to have been heard. Since then, during an evidence session before the House of Lords’ Communications and Digital Committee on 13 January 2026, the Secretaries of State Liz Kendall (Science, Innovation and Technology) and Lisa Nandy (Culture, Media and Sport) have said the government was wrong to have expressed any preference, and it is a “reset” moment.
All eyes are now on the government’s report on copyright and AI and economic impact assessment, which it must publish by 18 March 2026, to meet its copyright obligations under the Data (Use and Access) Act. Under the Act the government must consider each of the policy options put forward in the government's consultation on copyright and AI and make proposals, taking into account the consultation responses. For more on the government’s copyright obligations under the Data (Use and Access) Act, read our earlier article here.
Creative Content Exchange
During the 13 January evidence session, Liz Kendall and Lisa Nandy also mentioned the government’s plan to trial a Creative Content Exchange this year, to allow AI developers access to certain cultural content in return for fair remuneration. The Creative Content Exchange was first announced in June 2025 in the Creative Industries Sector Plan. Last month, as part of a wider announcement> on various government industrial strategy projects, the government said the trial will look at how the digitised content of certain institutions, including the National Archives, Historic England and various museums including the Science Museum and Natural History Museum, can be used by AI developers while respecting copyright. The pilot will test a range of commercial licensing models with the aim of launching an operational pilot platform by summer 2026.
Getty v Stability AI appeal
Later this year, we can also expect the outcome of the appeal in Getty Images v Stability AI, the first and only copyright case concerning AI in the UK to date. The highly anticipated first instance judgment was handed down in November 2025. The High Court held that Stability AI is not liable for copyright infringement arising from the download of copies of Stable Diffusion in the UK from online locations outside the UK (secondary infringement). Getty has since been granted permission to appeal. The appeal will focus on the meaning of ‘infringing copy’, for the purposes of secondary infringement (e.g. importation) in the context of an AI model.
The appeal will not determine whether training the AI model involves UK copyright infringement (because Getty dropped its case on primary infringement due to a lack of evidence of relevant acts in the UK). Nevertheless, the appeal has the potential to be very significant for both AI developers and rights holders. If the Court of Appeal find that the High Court reached the wrong conclusion on the ‘infringing copy’ issue, that would mean the importation of foreign-trained AI models (e.g. by downloading copies in the UK) would amount to UK copyright infringement, even if those models do not contain copies of the works on which they were trained but were merely exposed to those copies. However, this outcome seems unlikely.
For more detail on the judgment read our earlier here.
[View source.]
