The UK should introduce new legislation to control artificial intelligence or risk falling behind the EU and the US in setting the pace for regulating the technology, MPs have said.
Rishi Sunak’s government was urged to act as it prepares to host a global AI safety summit at Bletchley Park, home of the Enigma codebreakers, in November.
The science, innovation and technology committee said on Thursday the regulatory approach outlined in a recent government white paper risked falling behind others.
“The AI white paper should be welcomed as an initial effort to engage with this complex task, but its proposed approach is already risking falling behind the pace of development of AI,” the committee said in an interim report on AI governance. “This threat is made more acute by the efforts of other jurisdictions, principally the European Union and the United States, to set international standards.”
The EU, a trendsetter in tech regulation, is pushing ahead with the AI Act, while in the US the White House has published a blueprint for an AI bill of rights and the US senate majority leader, Chuck Schumer, has published a framework for developing AI regulations.
The committee report, whose introductory paragraph is written by the ChatGPT chatbot, lists 12 governance challenges for AI that it says must be addressed by policymakers and should guide the Bletchley summit, which will be attended by international governments, leading AI firms and researchers.
The technology has risen up the political agenda after breakthroughs in generative AI – the term for tools such as ChatGPT and Midjourney that are trained on vast troves of data taken from the internet – that can generate plausible text, image and audio content from human prompts.
The challenges include: addressing bias in AI systems; systems producing deepfake material that misrepresents someone’s behaviour and opinions; lack of access to the data and compute power needed to build AI systems; regulation of open-source AI, where the code behind an AI tool is made freely available to use and adapt; protecting the copyright of content used to build AI tools; and dealing with the potential for AI systems to create existential threats.
The government’s AI white paper published in March sets out five guiding principles for managing the technology: safety, transparency, fairness, accountability, and the ability of newcomers to challenge established players in AI.
The white paper said there was no intention to introduce new legislation to cover AI and instead expects different regulators – such as the UK data watchdog and the communications regulator, Ofcom – to thread those principles through their work, assisted by the government.
However, the document does refer to introducing a “statutory duty” on regulators to follow the principles. The committee report said that commitment alone pointed to the need for an AI bill in the king’s speech, which sets out the government’s legislative agenda for the next parliamentary year. Otherwise, the report said, “other jurisdictions will steal a march and the frameworks that they lay down may become the default even if they are less effective that what the UK can offer”.
If the government does not take action now, legislation may not be enacted until late 2025, nearly three years from the publication of the white paper, the report added.
The report also recommended that the safety summit should include as wide a range of countries as possible, amid speculation that China, a big tech and AI power, will not be invited.
Asked this week whether China should be invited to Bletchley Park, Greg Clark, the conservative chair of the committee, said: “If this is to be the first global AI summit then to have as many voices there as possible, I think would be beneficial. But it needs to be accompanied with a caveat that we don’t expect that some of the security aspects to be resolved at that level. Our recommendation would be that we need a more trusted forum for that.”
A government spokesperson said the potential of AI should be harnessed “safely and responsibly” and the forthcoming AI summit would address the threat of risks and harms from the technology. The spokesperson added that the white paper sets out a “proportionate and adaptable approach to regulation in the UK”.
The government has also established a foundation model taskforce, referring to the underlying technology for AI tools such as text or image generators, which will look at the safe development of AI models.
