The Joint Committee on Human Rights (JCHR) published a report calling for an AI bill that would tackle what it described as serious and potentially wide-ranging threats. The committee said the current regulatory approach is not fit for purpose and leaves gaps in protections.
Labour MP and committee chair Alex Sobel said there is no legislative and regulatory framework, including in the UK, that is adequate for AI. He warned that the technology is advancing with such speed and complexity that its impacts are difficult to anticipate, while the country remains unprepared for the consequences.
Risk-Based Regulation and Statutory Oversight
The JCHR said it wants a single, independent AI oversight body established on a statutory basis. It argued that the legal framework governing AI is fragmented and difficult to navigate, leaving some areas insufficiently covered.
The committee recommended that the proposed AI bill set up a regulatory regime that categorises different levels of risk. It said higher-risk AI systems and models should face more demanding obligations, covering responsibilities across the full AI lifecycle.
The report also called for obligations to be placed on those involved in designing, amending and using AI systems. It said responsibilities should be clear for participants in the supply chain and should ensure those duties are effectively addressed.
Concerns Including Bias and Misinformation
The committee highlighted concerns raised in recent years, including the potential for AI to develop bias against particular groups. It said systems trained on large datasets scraped from the internet could reflect racist, sexist and other harmful material.
It also pointed to the risk that AI could be used to create and spread misinformation.
In its 100-page report summary, the JCHR said AI has already been linked to abuses of human rights, giving examples including the creation of sexualised images of women and girls and the scanning of people’s faces without consent.
The committee said some uses of AI should be prohibited because they are incompatible with human rights. It cited areas for action including subliminal techniques, as well as inappropriate use of profiling and biometric data.
Government Position and US Proposals
Responding to the report, a government spokesperson said AI is advancing at an extraordinary pace and is creating both opportunities and risks. The spokesperson said the UK is working to test and understand AI, citing a security institute involved in developing standards and safeguards intended to keep people safe.
The JCHR’s publication comes as debate about AI regulation is intensifying internationally. On Saturday, the chief of US AI company Anthropic, Dario Amodei, set out a plan calling for regulation that he said should apply globally, include industry-wide rules, and involve independent monitoring of AI models as they are developed.
Amodei also urged development to be slowed down. He said the proposals had support from Sam Altman of OpenAI and Elon Musk, who described their alignment with the plan.
The report also references wider anxiety about AI. Jacob Coxon, an AI researcher who said he quit Anthropic, told the BBC that staff developing the systems were “genuinely frightened” about the future of humanity. An Anthropic spokesperson told BBC News the company is building models with what it described as some of the strongest safeguards in the industry.
Join the discussion?
14 September 2026
- 0 comments
- 29 views
-
webfact