UK Government Reverses Course on AI Copyright, Opens Door for Creative Industry Protection
The UK government abandoned its default position allowing tech companies to use copyrighted work without permission to train artificial intelligence systems. On 18 March, the government published a report on copyright and AI that reversed the approach embedded in the Data (Use and Access) Bill passed in early 2025.
Technology Secretary Liz Kendall said the government had "listened" to creatives, AI firms, and industry bodies. "We have engaged extensively with creatives, AI firms, industry bodies, unions, academics and AI adopters, and that engagement has shaped our approach," she said. "This is why we can confirm today that the government no longer has a preferred option."
The original bill required creators to opt out of AI training on a platform-by-platform basis rather than requiring tech companies to seek permission upfront. Artists and music producers including Paul McCartney and Elton John warned that the approach threatened the UK's £146 billion annual creative sector income and 2.4 million jobs.
Cambridge Model Shows Alternative Path
Cambridge University Press & Assessment (CUPA) has licensed its academic content for text and data mining for a decade. That model now sits at the centre of discussions about how to protect creative industries.
Catie Sheret, general counsel at CUPA, said the problem was not missing rights but "the practical and legal friction these differing rights create." She called for "a considered, fair, coherent approach based on the existing copyright frameworks in place around the world."
Mandy Hill, managing director at Cambridge University Press and president of the Publishers Association, said governments must "uphold the legal foundations of copyright when developing AI policies to avoid damage to the creative industries."
Collective Licensing Scheme Launches
Publishers' Licensing Services (PLS), a non-profit government-regulated organisation, launched a collective licensing initiative at the 2026 London Book Fair. The scheme allows publishers to opt into a shared licensing model for generative AI use.
The move addresses years of unauthorised use. Large language models developed by OpenAI, Google, and others were trained on copyrighted works without consultation or compensation to rights holders.
Dan Conway, CEO of the Publishers Association, said the government's shift was "a significant moment in cementing the government's reset on copyright and AI policy." He called for transparency legislation as a priority to support the UK's licensing market.
The reassessment process will take time, Kendall said, to "get this right." How that process unfolds will shape whether creatives can protect their work as AI systems become embedded across industries.
Your membership also unlocks: