What now for creatives, rightsholders, and AI businesses in the UK?

The UK government has published a coordinated set of responses addressing how to balance copyright protection with AI innovation. These include the Department for Science, Innovation and Technology’s (DSIT) Report on Copyright and Artificial Intelligence (“DSIT Report”), the Economic Impact Assessment, and the House of Lords Communications and Digital Committee’s report on AI and the creative industries.

AI could add approximately £55–140 billion to UK Gross Value Added (GVA) by 2030. The UK’s creative industries are world-leading; from Aardman animations to globally successful games such as Grand Theft Auto, they accounted for 13% of UK services exports in 2023 and are a central pillar of the government’s Industrial Strategy. Against that backdrop, the government’s response is cautious.

No immediate reform to AI and copyright law

The government has moved away from its preferred option of introducing a commercial text and data mining (TDM) exception with an opt-out mechanism for rights holders.

It concluded that stakeholder views were too divided and evidence too limited to justify legislative intervention to copyright law. It has also confirmed that it will not introduce a new regulator or impose additional regulatory duties at this stage.

This means that the existing framework under the Copyright, Designs and Patents Act 1988 (CDPA) continues to apply, including the narrow TDM exception limited to non-commercial research. The continued reliance on this framework leaves both AI developers and rights owners potentially exposed: AI developers face uncertainty around liability and future legislative changes, while creatives and rightsholders remain concerned about enforcement and their ability to monetise their works effectively.

The government has cautiously opted to monitor legislation and case law developments in the US and EU before committing to a definitive UK approach. This “wait and see” strategy leaves the UK without a clear short-term framework and may impact investment at a critical time for UK economic growth.

Licensing, transparency, and digital replicas

The DSIT report identifies three key areas for future focus:

1. Licensing and training data
   
   Licensing is expected to play a central role. The government has recognised that creatives and rights holders should be fairly remunerated if their works are used in AI systems but has stopped short of direct intervention. Market-led and sector-specific licensing models may emerge. Increasing demand from business users for “safe” AI tools is likely to also drive AI providers towards licensed datasets and stronger IP protections such as copyright shields.
2. Transparency
   
   Transparency remains under consideration, particularly in relation to training data and AI-generated outputs. Rather than imposing immediate statutory obligations, the government intends to monitor international frameworks and support the development of industry-led standards.
3. Digital replicas and deepfakes
   
   The treatment of “digital replicas” is expected to be an urgent area for reform. The government has recognised that unlike the US and parts of Europe, the existing UK law does not provide individuals with meaningful control over the use of their likeness, voice, or style, particularly in the context of deepfakes. A further consultation is expected in summer 2026, alongside ongoing work by Ofcom around the Online Safety Act.

Computer-generated works

The government has indicated a clear policy direction on the treatment of computer-generated works (CGWs). Under section 9(3) of the CDPA, where a literary, dramatic, musical, or artistic work is generated by a computer without a human author, the “author” is deemed to be the person who made the arrangements necessary for its creation.

The government has indicated that protection for CGWs should be removed, meaning outputs without sufficient human effort and skill would not be protected by copyright.

This potential change should not impact on AI-assisted works which involve human intellectual effort. Here, copyright will continue to subsist in the outputs (subject to any underlying third-party rights). The difficult balance will be the implementation of such a change; creative works are increasingly made up of aspects of CGW often generated by tools which have embedded AI features.

Economic trade-offs and international positioning

The economic impact assessment highlights the complex economic trade-offs at the heart of AI and copyright policy but provides limited quantitative certainty.

A more permissive regime for AI training data could support innovation and attract investment. However, weakening copyright protections risks undermining the UK’s creative industries due to uncertainties over licensing and enforcement.

A market-based approach to licensing may offer a pragmatic path forward. While larger players are already entering into licensing arrangements, smaller AI developers may struggle without access to clear, high-quality datasets. Equally, creatives will need confidence that they can negotiate fair terms and enforce them effectively.

Litigation

In the absence of legislative reform, the courts will play a critical role. UK proceedings such as Getty Images v Stability AI have already highlighted evidential and jurisdictional challenges, particularly where training takes place outside the UK. International cases, including Like Company v Google before the Court of Justice of the European Union, are expected to provide further clarity.

While they may shape the practical boundaries of lawful AI training in the near term, huge costly cases such as these do little to support clarity for SMEs and smaller rights holders.

Takeaways

For now, the UK remains in a period of uncertainty. Creatives, rightsholders, and AI businesses must operate within the current legal framework while monitoring developments closely.

AI developers should:

• Audit AI use and training data: review how training datasets are sourced and assess potential copyright and data risks.
• Contractual protection: review and cascade down any contractual terms from training datasets and consider copyright shields and indemnities for users.
• Develop or engage with licensing strategies: consider licensing frameworks with creatives; the government has signalled that market-led solutions are likely to play a central role. Users are increasingly likely to want to use AI tools that demonstrate they license copyright works responsibly and fairly.
• Be transparent: even in the absence of immediate regulation, AI businesses should anticipate greater expectations around disclosure of training data and AI-generated content.
• Monitor litigation and regulatory developments: ongoing cases in the UK, EU, and US will be critical in shaping how copyright law applies to AI training and outputs.
• Assess exposure to digital replica risks: organisations using generative AI should consider how outputs may replicate identifiable styles, voices, or likenesses and what safeguards they can and should consider applying here given this is an area likely to see reform in the UK.

Creatives and rightsholders should:

• Audit rights and content: identify key works, ownership and potential exposure to AI training, mining, or unauthorised use in AI tools.
• Strengthen contractual terms: ensure agreements and terms of use with publishers, platforms, and commissioners address AI use, including rights reservation and remuneration.
• Engage in licensing opportunities: consider how content can be licensed for AI training on appropriate commercial terms.
• Assert and protect rights: use available contractual and technical tools to reserve rights and monitor unauthorised use.
• Label works: ensure all works clearly contain copyright notices in all formats and contact information.
• Assess digital replica exposure: consider how likeness, voice, or artistic style may be used or replicated, and the extent of existing legal protection.
• Stay informed: review developments in law, policy, and market practice in the UK and internationally.

With the legal framework still evolving, early engagement with these issues will be critical. Those who act now will be better placed to manage risk and take advantage of emerging opportunities as the UK’s approach to AI and copyright continues to evolve.

If you have questions or concerns about AI and copyright, please contact Rebecca Steer.