Background
The proliferation of AI developers operating in the UK has generated numerous copyright concerns, which are compounded by the legal uncertainty behind the application of UK copyright law to the training of AI models.
On the one hand, AI developers face challenges in navigating copyright law in the UK, creating uncertainty which risks discouraging investment, innovation, and growth in the AI sector. On the other hand, rights holders, especially in the creative industries, find it increasingly difficult to monitor and control whether or not their works are being used to train AI models. In cases where works are not expressly licensed for AI model training, creators face a lack of compensation for such use. Therein lies the core debate: to what extent UK copyright law does, or should, restrict access to copyright-protected works for the purpose of AI training.
To help address these concerns, the Department for Science, Innovation and Technology (DSIT) launched a consultation (the “Consultation”) on copyright and AI which was presented to Parliament on 17 December 2024. The Consultation closed on 25 February 2025, and the Government has since been evaluating stakeholder responses. Those responses informed the Government’s subsequent March 2026 Report on Copyright and Artificial Intelligence (the “Report“), which provides further insight into the Government’s evolving position on AI and copyright. The Government will continue considering the evidence gathered through the Consultation and the Report to inform future policy.
Key Takeaways from the Consultation
The UK Government’s attempt to mitigate the legal uncertainty surrounding the application of copyright law to the training of AI models is a principal theme in the Consultation. The Government has stated that the overarching aim of the consultation process is to set out a plan to deliver a legal and regulatory copyright and AI framework that reconciles the need to reward human creativity, incentivise innovation in the AI sector, and provide the legal certainty necessary for long-term growth in both the creative and AI sectors.
Another objective is to promote greater trust and transparency between the creative and AI sectors, with the DSIT expressing doubt that maintaining the current copyright framework would achieve that end. Other options for shaping copyright and AI policy put forth by the DSIT include strengthening copyright laws by requiring licensing in all cases, a broad data mining exception, and a data mining exception under which right holders may reserve their rights, essentially acting as an opt-out option.
At the time of the Consultation, the Government preferred the opt-out data mining exception whereby AI developers would be able to train AI models using material to which they lawfully had access, but only to the extent that right holders had not expressly reserved their rights. However, since the Consultation closed, the Government appears to have backtracked from this approach. The proposal attracted significant criticism from creative industries and rights holders, who argued that an opt-out model would place an impractical burden on copyright owners to police and enforce their rights, while offering limited transparency regarding the datasets used to train AI systems.
Outcome of the Consultation
In its March 2026 Report, the Government cited strong views expressed in response to the Consultation (including the fact that 83% of respondents favoured stronger copyright laws), gaps in the available evidence, and the rapidly evolving AI sector and international context to explain why the opt-out data mining exception is no longer its preferred way forward.
Instead, the Government has outlined the following steps which it intends to take:
- Gathering further evidence on how copyright laws are impacting the development and deployment of AI across the UK economy;
- Considering and engaging with stakeholders on other potential policy approaches;
- Continuing to monitor developments in technology, litigation, international approaches, and the licensing market to inform its future approach to copyright and AI regulation.
The practical effect of these developments is that, despite the Government’s efforts to address copyright concerns, considerable legal uncertainty remains for both rights holders and AI developers, as there have been no material legislative reforms implementing the proposals discussed in the Consultation and the Report to date.
Implications and Next Steps for Businesses Operating in the UK
Nevertheless, what the Report does signal to AI developers operating in the UK is that they will no longer be able to rely on the prospect of a broad copyright exception with an opt-out as this is no longer the Government’s preferred approach following the Consultation.
Although the Government has not committed to a specific legislative solution and maintains that it does not currently intend to intervene in the licensing market, AI developers should remain alert to the prospect of a framework that requires licences for the use of copyrighted material in AI training. In anticipation of such a development, rights holders should therefore consider whether their existing licensing arrangements provide sufficient protection for their intellectual property assets.
The Report has also made it clear that the Government remains committed to increasing public visibility, particularly around “input transparency”, namely the disclosure of the material used to train AI systems, and “output labelling”, which refers to indicating whether outputs are generated by AI. Therefore, despite the absence of legislative intervention at this stage, drawing sensible inferences from the Report can help rights holders and AI developers prepare for the possibility of a firmer stance from the UK Government in the future.
While the Government has not yet committed to a specific legislative solution, the Report indicates a continued focus on protecting rights holders’ interests, improving transparency in AI development and gathering further evidence on the impact of AI training practices. Businesses should accordingly review their AI governance, data sourcing and licensing practices now, as a prudent risk-management measure. Taking proactive steps to assess training datasets, contractual rights and record-keeping processes can help organisations strengthen compliance readiness, reduce potential exposure to intellectual property disputes and respond more effectively to future regulatory developments.
Please contact Xavier Prida for detailed advice on the above.
The material contained in this article is only for general review of the topics covered and does not constitute any legal advice. No legal or business decision should be based on its content.
This article is written in the English language. Preiskel & Co LLP is not responsible for any translation of all or part of its content into any language.