Time to ACCCT: an AI copyright framework for UK creative industries
This report considers the challenges in developing a publicly available, machine-readable approach to copyright consent that protects the interests of copyright holders whilst allowing AI companies to legally access data. It is published by CoSTAR National Lab with the support of DECaDE and Sheridans.
Posted: 20 May 2025
Why we need an AI copyright framework for creative industries
Generative AI is rapidly reshaping the relationship between copyright, consent, control and access to creative data, raising urgent policy and infrastructure questions for the UK’s Creative Industries. In particular, the scale of automated data ingestion through web crawling makes traditional licensing, negotiation and enforcement models increasingly difficult to apply in practice.
The report proposed a scalable framework for responsible AI training and use of creative works, structured around five core priorities
— ACCCT: Access, Consent, Control, Compensation and Transparency.
It sets out a starting point for machine-readable consent infrastructure that could enable rights holders to permit or protect their work, while supporting lawful and transparent data access for AI developers.