Lord Chris Holmes: ‘It’s high time to legislate AI’

The author of the AI (Regulation) Bill spoke to Liz Hockley about why the UK needs to urgently establish its own approach to the technology.

UK government inaction on artificial intelligence (AI) is leading to creators’ works being used without remuneration, pushing businesses towards other jurisdictions and contradicting the country’s pro-innovation approach.

That is the view of Lord Chris Holmes of Richmond, a member of the House of Lords who advocates for the potential of technology and the benefits of diversity and inclusion.

Holmes’ private members’ bill, the Artificial Intelligence (Regulation) Bill, reflects his policy focus on digital technology for public good, and emphasises the importance of transparency, IP obligations and labelling.

He tells WIPR that the government’s wait-and-see attitude to AI has caused confusion and chaos—and that it’s “high time to legislate”.

‘Pressing need’ for regulation

Earlier this month, May 9, Holmes asked the government what steps it was taking to protect IP rights in relation to AI, since discontinuing plans to develop a code of practice.

He is particularly concerned about safeguarding the work of the creative industries—which he says provide some of the “sweetest elements of what it is to be human”—against its unauthorised use by AI developers to train large language models (LLMs).

Viscount Camrose, minister for AI and IP, responded in the House of Lords that modifications to UK copyright law would likely be part of the solution to balance AI innovation with IP rights.

Holmes said this tied in with the government’s long-stated position that the time to legislate for AI would come—but that time wasn’t now.

“The difficulty with their position is twofold,” he explains. “Firstly, they haven’t said what factors will need to be known that aren’t currently known, to enable them to move to legislation.

“Secondly, by saying they believe legislation is necessary but not saying when or how—that drives a whole load of uncertainty into the picture and also leaves creatives and IP rights owners having their works continually used without recompense, respect or remuneration.”

This creates an urgency for regulation, Holmes says. With a UK general election in July, he says there will be “a pressing need for the entire public, whichever lens you look at this through, for the King’s Speech to contain much on AI and indeed on other technologies”.

A fundamental truth

Referring to working group talks between rights owners and AI developers on a voluntary copyright and AI code of practice that broke down in February, Holmes says it was “desperately unfortunate when discussions fell apart without getting very far off the starting blocks”.

“Far more effort should be made to get the parties back around the table,” he contends.

Holmes stresses that creatives are not “luddites” and want to engage with AI, but in a way that is respectful and protective of their rights.

He believes the UK leadership needs to make it explicit that the ingestion of copyrighted material by LLMs constitutes infringement.

“The government and the UK Intellectual Property Office (IPO) could make a statement to make that fundamental truth clear to everybody across the economy and society,” he says. “No more information is required. No regulation. Nothing is required for them to make that point.

“It would be beneficial to the rights owners of course, but also to our society and the economy—and it would bring clarity to a situation where at the moment it’s confusion, it’s chaos, and it’s the taking of others’ hard work and endeavours without remuneration.

“It’s easily resolvable. It’s not as complex as is being made out.”

As well as harming the creative industries, Holmes believes that ambiguity in the UK around regulating AI is at odds with the government’s pro-innovation approach.

“For innovation, for inward investment you want certainty, safety, security,” he proposes. “It’s difficult to imagine people piling money in when you’ve been told that legislation will come but you don’t know when or how or in what form.”

This uncertainty could also lead businesses to look elsewhere and align themselves with jurisdictions that have legislated on AI, such as the EU with its AI Act.

And while Holmes describes the Act as a “valiant attempt” in all its 892 pages, he says it is overly prescriptive and complex—and, crucially, not the optimal outcome for the UK when it comes to legislating AI.

“We have close to a unique opportunity in the UK to do something with English common law that is respected and understood around the world—a stable, secure means of legislating,” he says.

A ‘light-touch’ AI Authority

Holmes firmly believes that right-sized regulation creates growth, is good for consumers, for innovation, for citizen rights, and drives investment.

“We all know bad regulation,” he tells WIPR. “It doesn’t mean because you’ve got bad regulation that regulation is bad.”

This “right-sized regulation” would be implemented by an AI Authority, according to Holmes’ bill, which he envisages as being “light-touch, agile, with a horizontal view”.

“The key thing is for people not to think this is going to be a huge, expensive, bureaucratic, do-it-all AI regulator,” he explains. “Not a bit of it. That wouldn’t achieve the outcomes we’re looking for.”

Firstly, the AI Authority would look across all existing regulators to assess their competency to address the challenges and opportunities presented by AI, and then identify the gaps.

It would also look at the statute book where relevant to consider the legislation’s suitability in terms of AI, and ultimately ensure a “consistent and coherent approach, which currently isn’t the case with the government’s well-intended though underpowered write-out to existing regulators”.

Untapped potential

Holmes puts forward that in many ways, the most important aspect of his AI Regulation Bill concerns public engagement.

“It’s so critical to get this right because no matter how good the technology, no matter how good the opportunity, if people don’t feel it is worthy of their trust then rightly and understandably they won’t engage with it,” he says.

This will mean that they won’t enjoy the potential benefits of that technology—but could be saddled with the burdens of it.

And engaging the public is, in his opinion, something we already know how to do and get right—for example, with breakthrough science such as in vitro fertilisation (IVF).

“What could be more [akin to] science fiction, more terrifying, than bringing human life into being in a laboratory test tube? Why is it not only seen as unthreatening but rightly seen as such a positive part of our society?” asks Holmes.

He attributes its success to the late Baroness Mary Warnock, who chaired a committee of inquiry that led to the creation of the Human Fertilisation and Embryology Authority (HFEA). This involved communicating to the public what the science meant, its impact, how it would operate and what the safeguards would be.

Holmes opines that we need a similar approach to AI if we are to appreciate its risks but also realise its opportunities for individuals, communities, cities and the country as a whole.

“We are capable of doing it but all of that currently largely remains potential. We need to make it inevitable.”