Source: IP Law Watch

The UK Government published a report following its copyright and AI consultation and set out likely scenarios for addressing the use of copyright-protected works to train AI models. The consultation focused on whether AI developers should be permitted to use such works for training without prior authorisation and, if so, under what conditions, with the Government having initially favoured a commercial text-and-data-mining exception paired with a rights-reservation mechanism and enhanced transparency obligations.

Responses revealed a clear divide: creative industry stakeholders overwhelmingly supported a licensing-first model stressing consent, control and remuneration, while the tech sector warned that mandatory licensing could hamper AI development and the UK’s competitiveness. The Government stepped back from a preferred opt-out licensing model in favour of further evidence gathering, technical work and monitoring of litigation and international developments. Key themes include proportionate transparency about training data sources and provenance of outputs, questions over protection for purely computer-generated works versus AI-assisted creations, labelling of AI content, and concerns about digital replicas and deepfakes in a jurisdiction without standalone personality rights. No immediate statutory changes were adopted; rights holders are advised to assess access and consider licensing or reservation mechanisms, and AI developers are urged to tighten governance and record-keeping and factor copyright risk into product design. The report underscores ongoing legal uncertainty, continued dialogue with industry, and a preference for alignment with international best practices. The blog piece is informational only and names Serena Totino and Simon Casinader as authors.