The Joint Committee on Human Rights, a committee drawn from both Houses of the United Kingdom Parliament, published Human Rights and the Regulation of AI as its Fourth Report of Session 2026 to 27 on Monday 14 September 2026. A select committee report carries no legal force. It sets out recommendations for the government to answer, and it marks the point at which pressure for United Kingdom AI legislation moves from commentary into a formal parliamentary recommendation.
The Committee calls on the government to bring in a dedicated AI Bill and to set up a new regulator. The design it recommends is risk based, classifying AI systems by risk level with more demanding obligations on higher risk systems, and it places accountability across the supply chain from design through deployment. It asks for certain uses to be prohibited as incompatible with human rights, naming subliminal techniques and inappropriate use of profiling or biometric data. It also asks for mandatory transparency so individuals know when AI is being used in decisions that affect them, for strengthened UK GDPR protection against misuse of automated decision-making, and for better routes to a remedy.
For firms deploying AI in the United Kingdom the report changes no existing duty, because a select committee recommendation binds nobody and sectoral regulators keep their remits. Its value to a deploying business is as a statement of where a future statutory duty would land: risk classification of each system, records that run back through the supply chain, and the ability to tell an affected individual that AI was used and to offer a route to challenge the outcome. Firms already holding EU AI Act documentation will find the recommended shape familiar.
The Committee records that the legal position is patchy and confused, with many regulators charged with tackling some AI harms but no single body coordinating regulation of AI. Its answer is one independent regulator holding statutory powers to set policy, monitor performance and enforce compliance. Whether the government accepts a single regulator rather than keeping AI supervision spread across existing bodies is the question to watch. The report also flags risk to already minoritised communities, including Black and Minority Ethnic people, which points at where early scrutiny would fall.
Licentium advises developers and deployers of AI systems on United Kingdom and European Union regulatory exposure and on preparing before a statutory duty arrives. Work we undertake includes risk classifying AI systems against the EU AI Act and the shape the Joint Committee recommends, building supply chain accountability records, reviewing automated decision-making against UK GDPR, and preparing responses to parliamentary and government consultations.