The Joint Committee on Human Rights (JCHR) has published its report on Human Rights and the Regulation of AI, calling for a dedicated AI Bill and the establishment of an independent AI oversight body.
The Committee’s report was published just as some of those at the forefront of developing AI were themselves arguing for the introduction of regulation, alongside a slowing down of the pace at which frontier AI labs improve the capabilities of their models. Whilst the conclusions reached by the Committee may not be what the heads of the AI labs had in mind – and it remains to be seen whether the Government will adopt any of its recommendations – the report provides an important indication of the level of concern about AI among UK legislators across the political spectrum, as well as the measures they consider necessary to mitigate its potential harms.
The report runs to just under 100 pages, covering everything from the meaning of AI and examples of the risks it poses to human rights, to discussion of the existing legal and policy framework.
On the latter point, the report is clear that “overall the legal landscape is patchy and confused. There are many regulators who are charged with tackling some AI harms, but no single body co-ordinates regulation of AI. This leaves gaps which mean people will be affected by AI harms, but will not be able to get redress“.
To address this, the Committee recommends establishing an independent AI oversight body. The body would act as a single point of contact for concerns about the use of AI, monitor emerging risks and harms, and maintain a repository of adverse incidents to help identify systemic problems.
In the Committee’s view, the oversight body would implement and enforce the provisions of a new AI Bill, including powers to:
- Publish mandatory codes of practice for the responsible use of AI in the public and private sectors, and sanction developers, deployers and other actors who fail to fulfil their responsibilities under those codes;
- Investigate alleged breaches of statutory prohibitions and mandatory codes of practice, including allegations that AI systems have caused harm to human rights; and
- Prescribe transparency requirements through its codes of practice, identifying what information must be disclosed by actors at different stages of the AI lifecycle, and to whom.
In addition to empowering the new oversight body, the Committee recommends that the AI Bill should prohibit uses of AI that are inconsistent with respect for human rights, including the development and provision of “very powerful AI systems (such as Artificial General Intelligence and Artificial Superintelligence) which risk causing widespread and very serious harm, including the capacity to evade effective human control”. At the same time, those systems that pose a high risk of causing harm to human rights would need prior approval before they can be deployed.
Finally, the Committee also recommends imposing “proportionate obligations” on actors at all stages of the AI lifecycle and supply chain to ensure that “risks are identified, evaluated and effectively addressed at the appropriate stage”.
To read the report in full, click here.