UK Government Should Clarify Whether It’s Investing in AI Companies that Use Unlicensed Training Data, Say Campaigners

UK Government Should Clarify Whether It’s Investing in AI Companies that Use Unlicensed Training Data, Say Campaigners

CMU (Complete Music Update)
CMU (Complete Music Update)Apr 29, 2026

Key Takeaways

  • Sovereign AI Unit controls a £500 million (~$635 million) national AI fund
  • Kidron demands public disclosure of the unit’s assessment criteria
  • Chair James Wise says compliance required, but no firm licensing rule stated
  • UK law lacks a commercial text‑and‑data‑mining exception
  • Unlicensed training data could undermine UK creative industries

Pulse Analysis

The UK government’s Sovereign AI Unit represents a landmark public‑private partnership, earmarking roughly $635 million to accelerate home‑grown artificial‑intelligence firms. By selecting an initial cohort of seven start‑ups and engaging with dozens more, the unit aims to keep AI expertise on British soil and signal confidence to global investors. However, the lack of transparent eligibility rules—especially regarding the use of copyrighted works for model training—has sparked criticism from the creative sector, which fears that public money could inadvertently subsidize infringement.

Copyright advocates argue that without explicit licensing mandates, AI developers may rely on loopholes such as foreign text‑and‑data‑mining exceptions or the U.S. fair‑use doctrine. The UK currently offers no commercial exception, and a previously proposed amendment was withdrawn after industry backlash. This regulatory gap creates uncertainty for music, film, and publishing stakeholders, who warn that unchecked data scraping could erode revenue streams and disincentivize new creation. The Getty vs. Stability AI dispute highlighted the difficulty of proving a legal duty to obtain licenses, underscoring the need for clear policy direction.

If the Sovereign AI Unit were to embed licensing requirements into its funding contracts, it could set a precedent for responsible AI development worldwide. Requiring licence‑cleared training data would not only protect creators but also foster a sustainable ecosystem where AI innovations complement, rather than cannibalise, the UK’s cultural industries. Transparent criteria would reassure investors, reduce legal risk, and demonstrate that public capital can be leveraged to promote both technological advancement and the long‑term health of the creative economy.

UK government should clarify whether it’s investing in AI companies that use unlicensed training data, say campaigners

Comments

Want to join the conversation?