Britain Has an AI Skills Shortage. Women Are Becoming a Smaller Share of the Workforce

Britain's artificial-intelligence labour market has two problems that are usually discussed separately. Companies say they cannot find enough of the skills they need, while women remain heavily underrepresented in the workforce supplying those skills. The government's AI Labour Market Survey 2025 shows why the two questions increasingly belong together.

Women accounted for 20% of AI roles in the survey, down four percentage points from 2020. At the same time, 97% of organisations surveyed identified at least one skills gap in the AI labour market and 57% reported a technical skills gap. More than a quarter said technical shortages had affected their ability to achieve business goals.

More companies employ women, but women's share is falling

There is one apparently positive signal. The proportion of surveyed organisations employing no women in AI roles fell from 53% in 2020 to 41% in 2025. In other words, female participation is spread across more employers than it was five years earlier.

But the overall share of AI jobs held by women still declined. That combination suggests the sector may be expanding faster than female participation. A company can therefore improve representation internally while the aggregate labour market becomes more male-dominated if new hiring and new roles grow faster elsewhere.

The shortage is most expensive at the experienced end

The AI labour survey identifies senior positions as particularly difficult to fill. That makes the representation gap more consequential. Experienced technical workers do not appear instantly when demand rises, and a narrow pipeline into senior roles can take years to correct.

International recruitment partly fills the gap, with 38% of businesses in the survey hiring from outside the UK. But employers also cited visa delays, costs and security-clearance constraints. Expanding domestic participation therefore sits alongside immigration and training policy as a practical source of additional capacity.

Pay inequality is a separate but related signal

The Office for National Statistics estimated the UK gender pay gap among full-time employees at 6.9% in April 2025, while the gap across all employees was 12.8%. ONS is explicit that this is not a measure of men and women being paid differently for the same job. It reflects the distribution of men and women across occupations, seniority, working patterns and other labour-market characteristics.

That distinction is particularly relevant to technology. If women are less represented in fast-growing technical roles and senior positions, occupational composition itself can contribute to wider earnings differences even without an employer paying two people differently for identical work.

Britain risks building its AI economy on a narrower talent base

The UK's AI strategy depends on turning investment into adoption across finance, healthcare, professional services, industry and the public sector. That requires more than frontier researchers. Businesses need engineers, product leaders, data specialists and managers capable of deploying AI safely inside existing organisations.

When 97% of surveyed organisations can identify a skills gap, underrepresentation stops looking like an issue that can be deferred until the labour market is less tight. The shortage is already here. Broadening participation is one of the few levers that can increase the domestic talent pool while also improving access to a high-growth part of the economy.

Our view: the 20% figure should be treated as an industrial warning

British Business Review's view is that the decline in women's share of AI roles is more important than a generic diversity score. Britain is trying to scale a strategically important industry while employers report that skills shortages are already limiting business performance.

The test for employers is not whether they can announce a target. It is whether recruitment, reskilling, return-to-work programmes and promotion systems measurably expand the number of women entering and progressing through technical AI careers. If the sector grows rapidly while representation falls, the UK will have expanded AI without expanding access to its opportunity at the same pace.