June AI policy update: the UK moves to measure and power its AI economy

As the UK invests billions in AI infrastructure, labs and adoption, a new government-backed institute sets out to answer the harder question: what impact is AI actually having on the economy?

By James Boyd-Wallis

UK updates

Treasury and DSIT launch AI economics institute

The Treasury and the Department for Science, Innovation and Technology (DSIT) have launched the AI Economics Institute, billed as the world’s first government-backed research institution dedicated to assessing AI’s economic impact.

The institute will weigh both the opportunities and risks under a range of scenarios, examining how AI is affecting productivity, labour markets and growth.

The launch comes at an important moment. Organisations including the Ada Lovelace Institute have argued for better projections of AI’s likely impact on productivity and the economy.

Jack Snape, DSIT’s deputy director of AI analysis, explains more in his post.

A £1.1bn plan to build Britain’s AI infrastructure

The UK government has also published a new £1.1bn AI hardware plan, setting out how it will back British companies developing the chips and semiconductor technologies behind AI.

The funding includes £750m for a new national AI supercomputer, with £400m assigned to new chips. In a bid to boost Britain’s sovereign AI capabilities, a further £150m will go towards buying novel chips from UK companies.

James Chalmers, DSIT’s next-generation hardware lead, provides more detail in a recent post.

Oxford and UCL to host new AI research labs

The government is backing two new AI research labs, led by the University of Oxford and UCL, to develop AI systems that are cheaper, more reliable and easier to run.

Oxford’s lab will focus on the science behind how AI systems learn. At the same time, UCL will develop models capable of running on more widely available hardware rather than relying solely on large data centres.

Oliver Purnell, partnerships lead at i.AI, argues the initiative should reduce dependence on the handful of companies that control frontier AI today and help build UK capability.

New coalition aims to accelerate AI adoption

A new AI adoption coalition, convened by industry bodies including techUK, has been launched to accelerate responsible AI adoption in the UK.

The three core targets are: to share knowledge and best practice across sectors; identify and address common barriers to adoption; and work with government to deliver on its AI ambitions. This new coalition aims to help businesses adopt AI and deliver a productivity dividend.

techUK’s Kir Nuthi highlights the initiative in her post, accompanied by an insight from Usman Ikhlaq.

Regulators focus on accountability, trust and following the law

At an All-Party Parliamentary Group on AI and Digital Regulation Cooperation Forum (DRCF) briefing in the House of Commons, three regulators set out how they are approaching AI.

Three key takeaways:

First, accountability: Charlotte Clark of the Financial Conduct Authority (FCA) described the regulator's ambition to drive the safe and fair adoption of AI, giving firms the confidence to adopt it through mechanisms such as sandboxes and live testing. She highlighted the Senior Managers Regime, ensuring responsibility ultimately rests with named individuals rather than AI systems, alongside the Consumer Duty, which requires firms to identify foreseeable harms and ensure consumers are not disadvantaged.

Second, trust: Karen Croxson of the Competition and Markets Authority (CMA) argued that consumer trust is critical to fostering AI adoption. The CMA recently put out a perspective on agentic AI and consumers, which raises fresh questions about whether agents perform reliably, whether they are faithful servants or can be manipulated, and what conditions will support a dynamic ecosystem. Companies should therefore be transparent about their use of AI, design up front for compliance, embed human judgement and step in quickly to address issues when they arise.

Finally, the time to act is now: William Malcolm of the Information Commissioner's Office (ICO) suggested that the time to get the questions around AI right is now, before the rules need unpicking later. We must learn from the past, and that is where the ICO is focused.

Open Rights Group calls for a digital policy reset

Ahead of Andy Burnham's succession to the premiership, the Open Rights Group set out what it wanted from a change at the top.

On AI, the organisation argues that cross-cutting regulation is a necessity as the technology is increasingly used in benefits, policing and borders. More broadly, it says a change in prime minister offers the chance of a digital policy reset.

Global update

Who is driving AI hype?

Christian Katzenbach, Vanessa Richter, Daria Dergacheva and Vasilisa Kuznetsova have published new research on who drove AI hype during the technology’s formative phase.

Drawing on 2010 data from Twitter (now X), the researchers argue that early narratives continue to shape today’s AI discourse and the regulation of generative AI. They identify notable differences between two countries.

In Germany, the discourse is strongly driven by actors beyond industry, such as political parties, NGOs and foundations. In the US, by contrast, AI discourse is characterised by economic dominance and individualisation.

James Boyd-Wallis is managing director of tech and AI-focused corporate and public affairs agency Highbury Communications, a Chartered PR practitioner, and co-founder of AI policy network Appraise.