Joint Committee Calls for Statutory AI Bill to Close Human Rights Gaps
The Joint Committee on Human Rights has called for a new, wide-ranging AI Bill to address the scale and seriousness of human rights threats posed by artificial intelligence. The committee published its findings on 14 September 2026. BBC News
The report runs to 100 pages. It is titled "Human Rights and the Regulation of AI" and is designated the committee's 4th Report. Committee chair Alex Sobel MP said "nowhere in the world, including the UK, has a current legislative and regulatory approach to AI that is fit for purpose."
Mr Sobel was elected Chair of the Joint Committee. The committee has 12 members drawn from both Houses of Parliament, currently including MPs and peers from the Labour, Conservative and Liberal Democrat parties.
Ministers said nothing yet in the verified material, but the committee's diagnosis was direct. The committee said the current legal framework applying to AI is fragmented and difficult to navigate, leaving gaps in protection.
Its prescription is primary legislation. The committee advocates creation of a single, independent AI oversight body on a statutory basis. It said the AI Bill should establish a regulatory regime classifying different risk levels and mandating more demanding obligations for higher-risk AI systems and models.
Some uses, the committee said, should be prohibited outright as incompatible with human rights. It cited potential action on subliminal techniques and inappropriate use of profiling or biometric data.
The committee cited AI-created sexualised images of women and girls and scanning people's faces without consent as examples of AI-related human rights abuses.
The inquiry also took evidence from industry figures. Anthropic head Dario Amodei proposed a plan including global regulation, industry-wide regulation and independent monitoring of AI models as they are developed, and urged slowing the pace of development. OpenAI boss Sam Altman and Elon Musk said they agreed with Mr Amodei's call to slow the pace of AI development.
Former Anthropic AI researcher Jacob Coxon told the BBC that staff developing AI systems were genuinely frightened for the future of humanity. An Anthropic spokesperson told BBC News the company was building models with some of the strongest safeguards in the industry.
Parliament's own announcement was titled "Wide-ranging AI Bill needed to address severe human rights risks posed by AI". UK Parliament The committee set 14 November 2026 as the due-by date for its Human Rights and the Regulation of AI inquiry. The inquiry will explore what regulation might be required to safeguard human rights when AI technologies are being developed and used.
The broader context here will be familiar to those who track AI policy across Whitehall and Brussels. EU governments and lawmakers agreed in May to delay rules on high-risk AI systems such as those involving biometrics. Reuters A UK upper-house committee said in March that Britain should back a licensing-first approach for AI training and should reject any move to let AI companies freely mine copyrighted material for commercial model training.
Looking at what this means for ministers and parliamentary business, the committee is asking for a different legislative model from sector-by-sector guidance. A statutory body with cross-cutting authority would cut across existing regulators, and a risk-tiered regime with outright prohibitions would require clear definitions in statute for subliminal techniques, profiling and biometrics. For government lawyers, the test would be enforceability and redress, not only principles.
For MPs and peers handling the bill if it comes forward, the cross-party composition gives the report procedural weight. Joint Committees speak for both Houses, and a unanimous call for prohibitions, tiered obligations and independent monitoring narrows the scope for a purely voluntary response. The next step, in parliamentary terms, is a government response and a decision on legislative time.
