In December 2024, the UK Intellectual Property Office proposed changes to existing UK copyright law that would permit AI platforms to permit data mining from copyrighted works without licenses. In its proposal, it noted that “There is debate about whether (the) balance within copyright law relating to the outputs of AI is right, between the traditional creator, the level of human input AI software might have, and the right holders in such works.”
Following the release of the proposal, the Office opened a 10 week consultation in December, which closed in February. It has yet to publish any follow-up or analysis.
Unlicensed use and its consequences:
“Some works are licensed to AI developers for the purpose of AI training. Others may be available under open licences. But in many cases, AI models are trained using works made available to the public on the Internet. These are often not expressly licensed for AI model training, and the creators of those works are not compensated for their use. There is limited disclosure about the sources of works used to train AI models and creators will often not know if their works form part of a training dataset.”
UKIPO proposed a multifaceted solution:
- Text and Data Mining (“TDM”) can be applied for any purpose, including commercially
- The party conducting TDM must have lawful access to copyrighted works, so rights holders can use legally-accepted means to obtain recognition and/or compensation
- The rights holder must pro-actively opt-out of TDM (and the proposal outlines several approaches to this)
- AI companies would be required to make disclosures about the materials used for training
Significant change
As current law (the UK’s Copyright Designs and Patents Act 1988 as amended in 2014) says that “(a) the copy is made in order that a person who has lawful access to the work may carry out a computational analysis of anything recorded in the work for the sole purpose of research for a non-commercial purpose, and (b)the copy is accompanied by a sufficient acknowledgement” and that “copyright in the work is infringed if … the copy is transferred to any other person, except where the transfer is authorised by the copyright owner;” then the 2024-25 proposal is a significant change indeed.
The law firm DLA Piper noted that “the (2024-25) Copyright and AI Consultation also goes beyond copyright and databases rights to touch on personality, privacy and performers rights… (and) also flags the importance of AI in education.” Current copyright law also did not anticipate the creation of deepfakes and computer-generated works by AI algorithms.
AI platform providers like the proposed exception
Predictably, AI platform providers took the side of open access. Google said that rights holders can already exercise control over copyright online and that opting out doesn’t guarantee protection. Google recommended that the UK maintain its current non-commercial research exception, enable TDM for any purpose, with rights reservation, and go ahead with the commercial exception.
OpenAI’s response positioned data access as funamental to “AI-driven economic growth and investment,” and took issue with implementation challenges and transparency obligations.
Media companies don’t like the exception
Speaking at the Enders/Deloitte Media and Telecoms conference in June, executives of the BBC and Sky both criticized the proposal to create the text and data mining exception to copyright law.
“As I look ahead to artificial intelligence, protecting copyright is a very big issue, and I think some of the consequences of the opt-out are impossible to police. If we as a large organisation spend the resource we do fighting for intellectual property rights, I can’t fathom how small producers keep up with a change of that nature. It is impossible to head in that direction,” said Sky Group CEO Dana Strong, as reported by The Guardian.
Doesn’t have all the answers
The UKIPO proposal did not purport to have the answers, but had made its suggestions. For example, while technical means exist to enable the automated expression of rights and usage – such as robots.txt – there are no standards; meaning that “right(s) holders often (have) to deal with multiple different systems. They are not always easily accessible, and may not meet their needs. There are also no requirements for AI firms to have such systems in place.”
Unlike the Office of the US Trade Representative, which posts the inputs collected for its own copyright-related reports, such as its Special 301 and Notorious Markets List, the UK does not publish the inputs that it receives from stakeholders.
One government voice responds
While the UKIPO has not yet issued its response to the inputs collected, the UK’s Information Commissioner’s Office, whose focus is more toward privacy and GDPR, issued its own resopnse in February 2025: Where a creator has not ‘opted-out’, their works could therefore be freely scraped, mined and used for the training of AI models. This proposal would bring the UK in line with the current approach in the EU pursuant to copyright laws…”
The law firm Ropes & Gray interpreted the ICO as saying, “(R)eliance on (the UKIPO’s proposed) new exception for commercial purposes where creators have not ‘opted-out’ should not be seen as a determination of a lawful basis for the processing of any personal data included in the mined data. The opt-out would allow the AI developers to use the data without infringing copyright law, but it would not constitute a lawful basis under the UK GDPR, which would still need be appropriately determined for any personal data contained therein.”
Further reading
Consultation: Copyright and Artificial Intelligence. Published December 17, 2024. UK Intellectual Property Office. Department for Science, Innovation & Technology. Department for Culture, Media & Sport
Section 29A, UK Copyright, Designs and Patents Act 1988. Accessed June 6, 2025. legislation.gov.uk
BBC and Sky bosses criticise plans to let AI firms use copyrighted material. Article. June 4, 2025. by Mark Sweney. The Guardian
The Information Commissioner’s response to the UK Government’s consultation on copyright and artificial intelligence. Report. February 25, 2025. UK Information Commissioner’s Office.
Training AI models: UK Government proposes EU style “opt out” copyright exception. Article. January 7, 2025. by Duncan Callow, Ally Clark. DLA Piper.
UK Government Consultation: Copyright, AI and the proposed ‘Opt-Out Model.” Article. December 17, 2024. Lewis Silkin.
ICO publishes response to the UK Government’s copyright and AI consultation. Article. March 18, 2025. by Catherine Keeling. Ropes & Gray LLP
Our response to the UK’s copyright consultation. Article. April 2, 2025. Open AI Global Affairs
Google’s pro-innovation proposals for the UK copyright framework. Article. April 2, 2025. Google
Why it matters
The question is basic: will the Government take the side of commercial interests by allowing data mining while weakening its copyright guard rails, or will it side with rights-holders whose creative works are a key to consumer targeting by parties whose main source of income is advertising?
The UK government’s proposal attempts to strike a balance between rights-holders and AI platform providers, and while it proposes protective practices for rights holders, it places a burden on those rights-holders to opt out.
Because most AI platforms are backed by large highly-resourced companies that exert great power in the market, Piracy Monitor believes that this burden is unreasonable; especially to smaller publishers that don’t have professional legal departments that can help them with copyright issues, or an many cases, the ability to afford legal counsel at all.
