Source: The Law Society Gazette

The House of Lords Communications and Digital Committee and the UK government have both examined how UK copyright law should respond to generative AI. The government has abandoned its previously preferred commercial Text and Data Mining (TDM) exception with a rightsholder opt‑out for AI training after scrutiny from creative industry representatives, and the Committee found the evidence did not show a TDM exception would significantly expand the AI sector while warning it could harm rightsholders and individual creators' livelihoods. The government will not make legislative changes at this stage and has said it will publish a final decision on its approach to AI and copyright law within the next 12 months; existing copyright law will remain in place and disputes will be resolved through the courts.

Generative AI can produce convincing digital replicas of a person’s likeness, voice or personality, creating risks when used without consent, and the UK currently lacks a standalone image or personality right, with protection coming from passing off, data protection, online safety regulation or defamation only in certain circumstances. Rightsholders report difficulty identifying whether their works were used to train AI models, which hinders enforcement and licensing negotiations; the Committee emphasised strengthening incentives to license works, transparency and enforcement within existing frameworks. The reports urge protections against unauthorised digital replicas and “in the style of” uses, note that labelling and mandatory disclosure of training sources are currently absent in UK law, and the government proposes working with industry on good practice for input transparency and labelling while not intervening in the licensing market for now.