The AI Adoption Plan for the Creative Industries marks an important change in emphasis. Rather than treating artificial intelligence as a future aspiration, it deals with the question how creators can use today’s technologies in practical and productive ways. The Plan adopts an augmentation-first approach, recognising that AI should support human creativity rather than replace it. If implemented well, it has the potential to encourage wider adoption while preserving the integrity of the UK’s creative industries.
The figures suggest that this transition is already underway. Around 51% of UK creative-industry businesses now use AI, although adoption varies considerably across sectors. IT and software lead at 60%, followed by design at 53%, film and television at 44%, while music and the performing arts remain much lower at 22%. This pattern is hardly surprising. Industries with mature digital workflows have been quicker to incorporate AI, whereas sectors built upon complex copyright relationships and licensing arrangements have understandably been more cautious.
The Plan’s eight recommendations reflect a pragmatic approach. They cover responsible AI, skills development, practical guidance, financial support and infrastructure. Much of the emphasis is on helping organisations develop confidence through training, shared resources, targeted SME support and expanded AI testbeds. As with much of the Government’s recent AI policy, however, the real challenge lies in maintaining an appropriate balance between encouraging innovation and protecting those whose livelihoods depend upon creative work. The Adoption Plan continues that balancing exercise and, admirably so.
The Plan reinforces the commitments set out in the Government’s Report and Impact Assessment of 18 March 2026. These include a consultation on digital replicas, the establishment of an AI labelling taskforce, a review of mechanisms enabling creators to exercise greater control over the online use of their works, and consideration of whether smaller organisations require additional support when licensing content. It also highlights the possible role of the Creative Content Exchange as a trusted marketplace for digitised creative and cultural assets.
That proposal deserves careful scrutiny. If properly designed, such a marketplace could improve licensing efficiency and support responsible AI development. Equally, if poorly implemented, it could inadvertently facilitate copyright infringement or evolve into a government-supported licensing platform that competes with creators and existing rights holders. The distinction is an important one. And a pilot using only public domain works for licensing falls dramatically short of the reality.
These initiatives now represent the core of UK policy. Yet the UK is only one participant in an increasingly competitive global AI environment. The European Union continues to implement its AI Act and associated regulatory framework. The United States remains committed to a more sector-specific approach, while countries including Saudi Arabia, China and India continue to invest heavily in AI capability. The UK’s policy choices therefore need to support both innovation and international competitiveness without weakening the legal protections upon which its creative economy depends.
For that reason, the wider policy discussion should remain focused on two issues that are fundamental to successful AI adoption.
1. Provenance and Transparency
The first is provenance. AI-assisted and AI-generated content should be accompanied by reliable information about the origin of training data, the ingestion process and the status of the resulting output. Without that transparency, creators cannot enforce their rights, rights holders cannot
monitor the use of their works, consumers cannot make informed choices, and markets cannot distinguish authentic creative works from synthetic ones.
The Government’s parallel work on AI labelling should therefore be viewed as far more than a technical exercise. Clear and reliable provenance is essential for market confidence. Charles Clark of the UK Publishers Association famously remarked during the early development of the internet that “the answer to the machine is in the machine.” That observation remains as relevant today as it was then. Technology itself must provide the mechanisms through which transparency and accountability are achieved.
2. Copyright as the Foundation of Creative Talent
The second issue is copyright. The Plan rightly recognises that copyright is not simply a legal doctrine but the economic foundation that enables creators to earn a living from their work. Its reference to respecting the original human creator under the heading of ethical AI is welcome, but this principle deserves greater prominence than a passing acknowledgement.
A workable system of copyright protection and enforcement remains the conditio sine qua non of responsible AI adoption. Questions surrounding training datasets, licensing, opt-out mechanisms, creator control and enforcement are not secondary matters to be resolved later. They determine whether AI becomes a tool that enhances human creativity or one that diminishes it.
Without credible licensing frameworks and effective enforcement, confidence in AI adoption will remain fragile regardless of how comprehensive the accompanying guidance or training may be.
Conclusion
The AI Adoption Plan is a constructive and practical document. Its recommendations on skills, guidance, infrastructure and responsible deployment should help creators engage with AI technologies more confidently. Nevertheless, the Plan cannot be considered in isolation from the Government’s broader copyright programme.
The principles identified in the 18 March Report remain decisive: no new copyright exceptions, effective labelling to distinguish human-created and AI-generated works, and meaningful transparency over the use of creative works, datasets and models during AI development.
Ultimately, successful AI adoption depends upon trust. That trust will not be created through guidance alone. It requires enforceable rights, transparent datasets, reliable provenance and licensing systems that command confidence from creators, technology developers and consumers alike. Without those foundations, the objectives of the Adoption Plan will remain difficult to achieve.