Summary

• UK GDP grew 0.4% in the second quarter, with information and communication among the strongest service sectors.
• Computer programming and consultancy output rose 3.7%, while scientific research and development also grew strongly.
• The pattern is consistent with continuing AI, software, cloud, and computing investment, but the national accounts cannot attribute the growth specifically to AI.

Britain’s technology investment cycle is leaving a clearer footprint in the national accounts, with programming, information services, research, and computing activity contributing strongly to second quarter growth even though the statistics cannot separate artificial intelligence from the wider digital economy.

The Office for National Statistics estimates that real GDP grew by 0.4% between April and June, following expansion of 0.6% in the first quarter. Services increased by 0.5%, construction by 0.3%, while production was broadly flat.

Information and communication output grew by 2.7% during the quarter, with computer programming, consultancy, and related activities rising 3.7%. Professional, scientific, and technical services increased by 1.7%, including 3.9% growth in scientific research and development.

Those figures arrive during heavy corporate and infrastructure spending on AI, software, cloud services, and computing capacity, although the ONS release does not describe the result as an AI boom or attempt to identify AI as the cause. Programming and consultancy cover a much wider range of work, while digital services had already been showing strength before the latest quarter.

AI sits inside a broader computing cycle

National economic statistics are poorly suited to isolating a technology that is embedded across existing industries. Spending on a generative AI project can appear as software, consultancy, cloud infrastructure, research, or hardware investment, while the company adopting the system may sit in banking, manufacturing, retail, government, or professional services.

That means an AI investment surge can become visible around the edges of several economic categories before statisticians can identify a distinct productivity contribution. Buying servers and cloud capacity raises technology spending immediately; reorganising a business process around those systems and producing more output with the same workforce can take considerably longer.

The current figures also reinforce an existing pattern rather than establishing a trend that began this quarter. Earlier ONS monthly data had already shown strong information and communication output, with computer programming and consultancy contributing materially, while companies have been increasing generative AI experimentation and cloud investment since well before 2026.

Computing strength therefore includes several overlapping forces. Businesses are still modernising software estates, moving workloads into cloud environments, increasing cyber security expenditure, improving data platforms, and building digital services, while AI is adding another large source of demand for infrastructure and specialist work.

Investment is easier to see than productivity

The stronger economic test will come later, when capital expenditure and professional services work should translate into higher output elsewhere in the economy. Britain has struggled with weak productivity growth for years, so a sustained increase in digital investment has value only if companies outside the technology sector eventually produce more, improve services, reduce errors, or use labour more effectively.

AI infrastructure also differs from many earlier software cycles because a substantial proportion of spending flows into physical assets. Datacentres require power, cooling, network connections, and construction, while advanced accelerators, memory, and other components are supplied through international manufacturing chains.

That gives the investment cycle a measurable domestic footprint where facilities are built and operated, although some spending also leaves the country through imported hardware and foreign owned platforms. The economic return consequently depends on more than the amount invested: ownership, supply chains, application development, skills, and the productivity achieved by organisations using the infrastructure all influence how much value remains in Britain.

Early GDP estimates are also revised as more information becomes available, which makes a single quarter an unreliable foundation for claims about a structural shift. Technology subsectors can be volatile, and the UK’s overall expansion was spread across several industries rather than being created by computing alone.

Even with those limits, the latest figures provide firmer evidence than another adoption survey or vendor forecast. Computing intensive activity is contributing meaningfully to growth, and the next stage is to see whether that investment spreads into stronger productivity across the sectors paying for the technology rather than remaining concentrated among the companies building and selling it.