By Thomas Lacy, IFOW Policy and Public Affairs Lead & Dr Clementine Collett, BRAID UK Research Fellow at University of Cambridge and IFOW Research Fellow

Introduction

The explosion of generative AI (GenAI) into society has thrown up critical questions for UK creative sectors - relating to copyright, licensing, and fair renumeration for creative works which have been used to train GenAI models.

On the eve of upcoming, legally required Government assessments on policy in this area, and with continuing creative sector disquiet showing no sign of abating, we reflect on the current policy landscape for AI and the creative sector, how we got here, and what needs to come next to preserve, protect, and enable the creative industries to thrive in a changing technological environment.

The disruptive impact of GenAI on creatives is at this point well established

The Impact of Generative AI on the Novel report, co-published by IFOW and the Minderoo Centre for Technology and Democracy at the University of Cambridge, showed that 39% of novelists report that GenAI has had a negative impact on their income, and 51% of novelists think they will be replaced by AI entirely.

The ‘Brave New World?’ report, published through a collaboration between The Society of Authors, Association of Illustrators, Independent Society of Musicians, The Association of Photographers, and Equity, contains equally stark findings, including that almost a third of illustrators, more than half of photographers, and almost three quarters of musicians see their see their income negatively affected by AI to tune of several thousands of pounds.

AI’s impact also extends to the structures of how we access content. For instance, at a recent Copyright, AI, and the Creative Industries conference convened by Westminster Insight, attended by this blog’s authors, Sajeeda Merali of the Professional Publishers Association clearly laid out how AI summaries – those embedded in search engines or carried out using AI programmes – capture value by taking information from online publications without directing traffic to their sites.

In the UK, as in many other countries, these factors have sparked intense debate – for well over a year now – and policymakers and courts around the world continue to grapple with the ethical, legal, and regulatory challenges these technologies pose.

Framing the UK Debate

One of the key policy issues in the UK debate relates to how creative works are used to train GenAI: both the AI and the Novel and Brave New World reports found most creatives believe their work has been scraped to train generative AI without their consent, and many creatives are calling for better protection of their works.

Much of the current disquiet stems from the Copyright and Artificial Intelligence consultation, first published in December 2024, which saw the Government state a preference to implement an exception to copyright law with a ‘rights reservation’. This would allow for AI companies to use creators’ works to train AI models, unless those creators had explicitly reserved their rights.  This has become known as the ‘opt out’ model,  and has faced widespread criticism as a move to prioritise technology companies’ access to data over the rightful control that creatives’ ought to have over their work.

This proposed model caused huge backlash from the creative sector. Intentionally or not, in this move the Government framed the creative and AI industries as being pitted against one another, when in reality, research has shown that creatives are early adopters of technology and simply want to be asked for permission, and given payment, for the use of their work.

The Waiting Game

Thanks to Baroness Kidron’s campaign to amend the Data (Use and Access) Act, the Government is legally bound to produce an economic impact assessment and a report on each option set out in the consultation, by March 2026. We are also waiting to see what happens with developments in licensing, including the creation of the Creative Content Exchange (CCE) and the CLA/ALCS license. Meanwhile, we are also beginning to see individual licensing deals being struck in the market, and a series of legal cases in various jurisdictions testing the limits of copyright law.

Amidst this the Government is conducting working groups, the results of which are not yet clear. At the recent conference, some participants told us that these groups have been stuck on basic educational points around copyright, with some in Government (according to one participant) still describing copyright as a barrier to be overcome.

The question is whether this waiting time for development and information gathering will lead to practical steps or not, particularly for small players seeking to engage with licencing in a realistic and practical way, and for creatives who need to be able to access these benefits. The Government needs to reach considered but actionable steps soon to provide clarity to AI companies and creatives alike, particularly as uncertainty could be holding back licensing deals.

Dispelling Myths

At the recent conference, we heard about a few myths which need to be dispelled around copyright and AI. Primarily, speakers discussed the myth that unless copyright law is relaxed to fit with international contexts, this will harm UK innovation. Well-trodden comparisons to the US, Japan, the EU, and Singapore were discussed. However, participants pointed out that these legal contexts are not as relaxed as they are often made out to be, and that none of them offer a clear-cut example of what the UK should be aiming to replicate.

Participants told us that in the US we are still seeing over 100 ongoing legal cases around AI and copyright. We heard that in Singapore, despite relatively permissive rules, more AI growth is still seen in neighbouring Johor Bahru in Malaysia. Japan also has many limits on its Text and Data Mining (TDM) rules, and countries like Australia have turned away from a TDM exemption altogether.

Meanwhile the UK, with its robust copyright laws, remains third in the world for AI investment, behind the much larger US and China. We must refute the myth that copyright law is hindering innovation. Overall, the UK is not an outlier in its copyright regime, and the idea that copyright law could only be tweaked in order to cause a flood of innovation is not tested or evidenced.

Making the Future Work

Copyright is one of many challenges faced by creative workers in the context of technological change. IFOW’s CREAATIF project explored wider impacts at the societal and ecosystem level, on social relationships and interaction, and on individual creative workers. Key findings on matters ranged from shifting skills portfolios for creative workers, challenges in access to good work, to challenges around creative novelty in outputs.

The impact of AI on the creative sector will be one theme at IFOW’s upcoming Making the Future Work Conference in London,  on18 May 2026 – a conference bringing together senior decision-makers, expert academics and leading firms to share best practice, reflect on current approaches and envision new futures for the overarching question: how AI can be coupled with human capability to turn the future towards good?