UK government puts brakes on opt-out copyright exemption for AI

The UK government has said it will not implement its plan to allow artificial intelligence (AI) firms to use copyrighted material without explicit permission for now, but has not ruled out some form of copyright exemption down the line.

In late 2024, the government proposed allowing AI companies to train their models on copyrighted works unless rights holders explicitly opted out, meaning that rather than requiring AI companies to seek permission from rights holders for the use of their work, the burden would be placed on the creators themselves to actively object.

The opt-out proposal – which the government said was its preferred option at the time – provoked significant backlash from creatives, who viewed it as too conciliatory to the narrow interests of tech companies.

Of the more than 10,000 people who responded to the government’s previous consultation on how AI is affecting copyright, just 3% backed its opt-out proposal.

Following the closure of that consultation in February 2025, the government has now published a progress report and impact assessment, which outlines its current thinking on the four options proposed.

While technology secretary Liz Kendall said on 18 March 2026 that the government “no longer has a preferred option” on copyright reform after the consultation backlash against the opt-out proposal, this was already confirmed as early as May 2025 by sources close to then technology secretary Peter Kyle, who made clear the proposal was one of several now being considered.

The four options being considered include keeping copyright and related laws as they are; strengthening copyright to require licences in all cases; implementing a broad data mining exemption for AI companies; or creating a more limited data mining exemption that allows copyright holders to opt out, underpinned by measures to promote and support greater transparency from developers.

While the government has now formally ruled out a broad copyright exception with no opt-out for creatives, it makes clear that some form of copyright exemption is still on the table, and has only dropped the opt-out proposal as its “preferred” option.

“We said we would not take forward this approach unless we can be confident that it will meet our objectives,” it said. “But there are still significant gaps in the evidence base – including in terms of how copyright impacts the wider development and deployment of AI – and uncertainty about the impact of reform.

“In view of the concerns raised by stakeholders, and the continued uncertainty about the likely effects of an exception with opt-out, a broad copyright exception with opt-out is no longer the government’s preferred way forward.”

Elsewhere, it added that while continuing to build up the evidence base, the government will continue to monitor the wider market and international regulatory developments in this area, while also considering alternative options.

The government said that while this will include considering “focused” copyright law exemptions on, for example, research and public interest uses, these approaches need “further consideration”.

Although the government has made no firm conclusion on what will happen next, it said it would not reform copyright laws "until we are confident that they will meet our objectives for the economy and UK citizens".

Commenting on the development, Ed Newton-Rex, a prominent commentator on AI and intellectual property, said that fundamentally, it represents nothing more than a delay: “The government is simply declining to make a decision on AI and copyright yet, when one is sorely needed. Every day that this delay continues is another day that the UK is set back.

“Only when the government commits to standing by existing copyright law will the market for licensing training data to AI developers be able to reach its full potential – and only then can the UK get on with the important work of building a responsible AI ecosystem based on respect and fairness rather than on exploitation.”

Mandy Hill, managing director of Cambridge University Press and the president of the Publishers Association, said while the government’s backtrack was a victory “over the self-interest of a handful of large corporations”, it has not entirely ruled out allowing AI developers to use copyrighted material to train their models without a licence.

“The existing law is clear,” she said. “Copyright material cannot be used for AI development and training without permission.”

Jo Twist, chief executive of trade body for the UK music industry, the BPI, added that while she was “pleased” the government has formally changed its position, “we don’t want to see any kind of text and data mining exception being introduced further down the line”.

Antony Walker, deputy chief executive at trade association TechUK, said the announcement must be used as an opportunity to “reset and find a way forward” on copyright.

“The UK has set its sights on leading the G7 in AI adoption, but that requires a clear and enabling framework for AI innovation,” he said. “With international competitors moving ahead, the UK cannot afford for this to remain unresolved.”

UK Music boss Tom Kiehl said the hundreds of thousands of people in the industry should be able to work “without the constant fear that the fruits of their labour could effectively be taken by AI firms without payment or permission”.

The Musicians’ Union, meanwhile, has called for collective licensing schemes to protect individual artists, “not just major rights holders”.