The UK government has stepped back from one of its most controversial proposals on artificial intelligence and copyright, signalling a decisive shift towards market-led licensing and greater transparency rather than sweeping legal reform.
In its long-awaited Report on Copyright and Artificial Intelligence, published in March 2026, ministers confirm they will no longer pursue a broad copyright exception for AI training with an opt-out mechanism — a policy that had triggered fierce opposition from across the UK’s creative industries.
Instead, the government is opting for a more cautious, evidence-led approach, prioritising transparency obligations and allowing a nascent but rapidly expanding licensing market to develop. The move marks a significant recalibration of policy at a time when the UK is seeking to position itself as both an AI superpower and a global creative hub.
At the heart of the report is a clear admission: the government’s preferred option, allowing AI developers to use copyrighted material unless rightsholders explicitly opted out, failed to win support.
The consultation attracted more than 11,500 responses, with the overwhelming majority of creators, publishers and rights organisations rejecting the proposal outright.
Ministers now concede that a broad exception “with opt-out is no longer the government’s preferred way forward”, citing strong industry opposition, lack of consensus, and insufficient evidence on economic impact.
This represents a notable victory for the UK’s creative sectors, from publishing and music to film and photography, which argued that such an exception would effectively legalise uncompensated use of their work by generative AI systems.
The report lays bare the fundamental policy dilemma: how to balance AI-driven economic growth with the protection of intellectual property.
On one side sit AI developers, who require vast datasets, often including copyrighted material, to train large language models and generative systems. On the other are creators whose works underpin those systems but risk being displaced by them.
The government acknowledges that modern AI models are typically trained on “billions of copyright works”, raising complex questions about fairness, consent and competition.
Yet it also highlights uncertainty around the economic benefits of reform, noting limited evidence that loosening copyright rules would materially increase AI investment in the UK.
In effect, ministers are choosing to pause rather than gamble.
Rather than legislating, the government is placing its bets on licensing, a market-based mechanism already beginning to take shape.
A growing number of deals between AI firms and content owners, particularly in publishing, music and image libraries, suggests a commercial model is emerging. However, the report acknowledges this market is still “new and evolving” and lacks transparency.
Crucially, ministers have ruled out direct intervention for now:
“We propose not to intervene in the licensing market at this stage… and will keep market-led approaches under review.”
This position aligns closely with industry sentiment across both creative and technology sectors, which broadly favour voluntary, negotiated agreements over statutory schemes.
However, it also raises important questions, particularly for SMEs and individual creators, about bargaining power and equitable remuneration.
Among those welcoming the shift is Tom West, CEO of Publishers’ Licensing Services (PLS), who sees licensing as both practical and scalable.
West said: “We welcome that the government has listened to the strong response it received from across the UK’s creative industries to its consultation and has stepped back from its preferred option of a copyright exception with an opt out and is to review the transparency of AI inputs, which would further boost licensing.
Whilst we await further clarity from the government on the long-term direction of its copyright policy, PLS will continue to serve our publishers and work with our partners on market-based, industry-backed AI licensing solutions.
This approach is already being put into practice. At the London Book Fair last week, PLS launched the first stage of a new collective licensing solution designed specifically to support the use of published content in AI. It was met with strong interest and positive feedback from publishers and industry partners, with publishers already beginning to sign up. The solution offers a practical, scalable way for AI developers to access high-quality content while ensuring creators are paid and retain control over how their work is used.
The case has not been made for the introduction of a new copyright exception. There is no market failure and a dynamic licensing market for the use of content in AI has developed and continues to grow. Any copyright exception for generative AI would jeopardise these licensing solutions, removing the ability of large and small rightsholders to receive payment for the use of their works in AI and reducing control over their content.
PLS welcomes the government’s engagement on this critical issue. We share a commitment to a mutually beneficial outcome and invite the government to work closely with us to help further develop and promote licensing options that support rightsholders of all sizes and AI developers seeking high-quality, trusted content.”
If licensing is the economic mechanism, transparency is the regulatory lever.
More than 90% of consultation respondents supported requirements for AI developers to disclose the sources of training data.
The government agrees, in principle, but stops short of immediate regulation. Instead, it proposes:
• developing industry-led best practice
• monitoring international frameworks (notably the EU AI Act)
• considering future legislation if needed
Transparency is seen as essential to enable enforcement, licensing and trust, particularly given that creators often have no visibility over whether their work has been used.
For UK businesses, particularly SMEs, the implications are nuanced.
For creators and publishers
• greater protection in the short term
• stronger negotiating position in licensing deals
• ongoing challenges around enforcement and visibility
For AI startups and developers
• continued legal uncertainty
• potential cost barriers to accessing training data
• reliance on licensed or overseas-trained models
For the wider economy
• slower regulatory clarity
• reduced risk of over-regulation
• continued dependence on global AI ecosystems
The report explicitly notes that SMEs on both sides, creators and developers, face disproportionate challenges under the current system.
Perhaps the most striking aspect of the report is its tone: cautious, iterative, and deliberately non-committal.
The government repeatedly emphasises the need for more evidence, more international alignment, and more market development before taking decisive legislative action.
With ongoing litigation in the US, new rules emerging in the EU, and rapid advances in generative AI, the UK risks being pulled in multiple directions, economically, legally and politically.
This is not a resolution, it is a holding position.
By stepping back from sweeping reform, the government has bought time. But it has also shifted responsibility onto the market to prove that licensing can work at scale, fairly and efficiently.
If it can, the UK may yet carve out a balanced model that supports both innovation and creativity.
If it cannot, the debate over copyright and AI will return, sharper, louder, and far harder to resolve.
Read more:
UK government backs away from AI copyright overhaul as licensing emerges as the battleground