A cross-party group of UK MPs and peers is calling for a dedicated artificial intelligence law, warning that existing protections cannot keep pace with rapidly developing technology. The Joint Committee on Human Rights says unchecked AI threatens privacy, equality and access to justice, while ministers argue that regulation is already advancing.
Key takeaways
- The Joint Committee on Human Rights says current UK AI rules are fragmented and inadequate.
- It wants a new law covering the entire AI supply chain and technology lifecycle.
- Some uses, including harmful profiling and certain biometric applications, could be banned.
- The government says action is already being taken but has not introduced a dedicated AI bill.
MPs call for a single AI regulator
The committee’s 100-page report argues that the UK needs a statutory, independent body with power to set policy, monitor AI systems and enforce compliance. At present, responsibility is spread across existing regulators, creating what the committee describes as gaps and uncertainty.
Committee chair Alex Sobel said no country, including the UK, currently has an AI regulatory framework that is “fit for purpose”. The proposed legislation would apply stronger duties to higher-risk systems, from development and testing to deployment and retirement.
For Power AI Media readers tracking the fast-moving policy debate, the recommendation marks a shift from voluntary safeguards towards enforceable oversight across the AI ecosystem.
Privacy, discrimination and access to justice
The report identifies three central human rights concerns: equality and non-discrimination, the right to an effective remedy, and privacy and data protection.
AI systems trained on large internet datasets can reproduce racial, sexist or other forms of bias. Facial recognition and other biometric tools may also be used without meaningful consent. The committee pointed to sexualised images created without permission as an example of how generative AI can cause direct harm, particularly to women and girls.
It also warned that people affected by automated decisions may struggle to understand how those decisions were made or challenge them effectively. That concern extends across areas covered by Power AI Media, including healthcare, cybersecurity, employment, content generation and public services.
Calls for prohibited uses
The committee says some applications should be prohibited outright where they are incompatible with human rights. Possible examples include subliminal techniques and inappropriate uses of profiling or biometric data.
It is also urging tougher transparency and accountability requirements for high-risk models. These could include stronger testing, documentation, impact assessments and routes for people to seek redress when an AI system causes harm.
Government faces pressure over promised legislation
Labour’s 2024 election manifesto promised binding rules for the companies developing the most powerful AI models. The government also committed in the 2024 King’s Speech to introduce appropriate legislation, but no dedicated AI bill appeared in the latest programme.
Education minister Georgia Gould said ministers were already acting, citing measures on non-consensual sexualised deepfakes and cybersecurity. She also highlighted the AI Security Institute’s work testing models and assessing future risks, while stressing the need for international cooperation.
Industry split over the pace of development
The parliamentary report comes amid a wider argument about whether AI development should slow. Anthropic chief executive Dario Amodei has called for independent monitoring, international coordination and a slower pace of progress, warning that self-improving systems could outstrip society’s ability to control them.
OpenAI’s Sam Altman and xAI founder Elon Musk expressed support for pacing frontier AI. US President Donald Trump, however, rejected calls for a slowdown, arguing that strong political leadership and existing criminal and regulatory powers were sufficient.
The dispute underlines the challenge facing policymakers: encouraging useful innovation while ensuring that privacy, safety and basic rights are not treated as afterthoughts.