UK MPs and peers call for an AI Bill and a single regulator; government reply due by mid-November

The United Kingdom still has no dedicated AI law, and a cross-party committee of both Houses of Parliament wants that to change. On September 14, the Joint Committee on Human Rights (JCHR) published its report Human Rights and the Regulation of AI. It concludes that current UK rules are not fit for purpose. Under parliamentary convention the government has two months to reply, which points to mid-November.

What the committee recommends

The report argues that today's framework is fragmented across sectors, focuses mainly on those who use AI rather than those who design it, and leaves regulators unable to test systems before release. Its main proposals are:

• an AI Bill using a risk-based approach, with tougher duties for higher-risk systems and models;
• bans on uses it sees as incompatible with human rights, such as subliminal techniques, emotion inference and some forms of profiling or biometric use, subject to public consultation;
• prior approval by a public body before high-risk systems are deployed, plus due-diligence duties across the AI supply chain;
• transparency duties, including telling people when AI is used in decisions that significantly affect them;
• stronger UK GDPR rules on automated decisions, so that a nominal human in the loop is not enough;
• a single independent AI regulator with powers to issue codes and sanction wrongdoing, and a statutory footing for the AI Security Institute, which now relies on voluntary arrangements with developers.

Committee chair Alex Sobel, a Labour MP, said the UK is currently "unprepared to deal with its consequences".

Where the government stands

The committee notes that the government promised AI legislation in the 2024 King's Speech but has not yet introduced a bill covering the most powerful models. The UK has signed the Council of Europe Framework Convention on AI but has not ratified it, and the report asks for a ratification timetable. Not everyone sees a gap: A&O Shearman reports that the Information Commissioner's Office told the committee that AI's privacy challenges are not insurmountable under current law.

What to watch

The government's response will show whether ministers are ready to commit to a bill, a timetable or a new regulator. They are not obliged to accept any recommendation. The report adds to pressure on the UK's lighter, sector-led approach at a time when the EU AI Act is being phased in and South Korea's AI Basic Act is already in force.