Oxford Report Calls for Tougher AI Copyright Rules to Protect UK Creators

The UK could become a global leader in licensing creative work for artificial intelligence, but the Government must strengthen copyright protections and establish clearer rules for AI companies, according to a new report from the University of Oxford.

Published on 13 October, the Oxford Consultation on Copyright and AI brings together contributions from more than 40 organisations and individuals across the creative industries, technology, law, policy and academia. It sets out a framework intended to let AI developers access high-quality material legally while giving creators greater control over how their work is used and ensuring they can be paid.

The report was launched at the Creators Emergency Summit in London, where artists, campaigners and policymakers gathered to discuss the impact of generative AI on creative livelihoods.

Three changes proposed for AI copyright

The report identifies three priorities for reform: ownership and consent, transparency and traceability, and fair licensing.

First, creators need greater control over their work. The report argues that copyright holders should be able to make meaningful decisions about whether and how their material is used by AI systems. This follows a prolonged dispute over whether copyrighted work should be available for AI training unless its owner actively opts out.

The UK Government previously faced significant opposition to an approach based on opt-out provisions, and ultimately abandoned its preferred option during the passage of the Data (Use and Access) Bill.

Second, AI companies need to provide greater transparency. Without information about the material used to develop AI models, creators may struggle to establish whether their work has been included or to exercise their rights. The report calls for stronger transparency, traceability and disclosure so that the use of protected material can be better understood.

Third, the UK needs a functioning licensing market. Clear rules and workable licensing arrangements could allow AI developers to obtain creative material legally while giving rights holders a route to payment. The report argues that stronger copyright protection need not prevent technological progress and could instead help establish a competitive market for licensed training data.

Why this matters to the creative industries

The proposals affect far more than visual artists. Writers, musicians, photographers, performers and other creative professionals face questions about whether their work is being used to develop commercial AI systems, what control they have over that use and whether they receive compensation.

Baroness Beeban Kidron, who wrote the report’s preface, warned that creators need to present a united case for stronger protections. She argued that payment and control over creative work risk being taken away if the industry fails to secure meaningful change.

The report also identifies an economic opportunity for the UK. Its established creative industries, alongside its expertise in technology, law and regulation, could help the country develop a trusted market for licensed material used in AI development.

Will the Government act?

The report is a set of recommendations, not a new law or a licensing system that has already been introduced. Implementing its proposals would require decisions about copyright rules, disclosure requirements and how licensing arrangements would operate in practice.

Its central argument is that the UK should not have to choose between supporting AI development and protecting the people who create the material that helps power it.

For creators, the proposed changes could provide a clearer route to exercising their rights and receiving payment. For AI developers, a well-defined licensing framework could offer greater certainty about how creative material can be used legally.

The report now adds a specific set of proposals to the ongoing debate over AI and copyright, putting pressure on policymakers to determine how those principles could be translated into practical rules.