UK CFOs Expect AI Job Cuts—but Only 6% of Adopters Report Them

In the UK, Deloitte’s 14 October 2026 CFO survey found that 52% of 66 finance chiefs at major companies expect AI to reduce headcount in their own businesses over the next 12 months. That is a forecast of what they think will happen, rather than a count of jobs already lost.

The Office for National Statistics’ 8 October 2026 bulletin gives the latest reported outcome: in September 2026, 6% of AI-using businesses with at least 10 employees said adoption had reduced their overall headcount. The surveys cover different businesses and different points in time. Together, they show why an expectation among large-company leaders cannot yet be read as widespread displacement across UK employers.

What the finance chiefs expect

The CFO poll ran from 17 September to 1 October. Its respondents included finance chiefs from 12 FTSE 100 companies and 25 FTSE 250 companies; the remainder worked for other listed firms, large private companies and UK subsidiaries of major overseas-listed businesses. It captures the outlook of substantial corporate employers, rather than a representative count of all UK firms or workers.

Headcount was one of several effects respondents anticipated from AI over the coming year. In the same poll, 48% expected lower operating costs and 37% expected higher margins. Cost reduction was a strong business priority for 68% of respondents, up from 53% in the previous quarter. Those answers put the staffing forecast in the context of wider pressure to improve efficiency, but they do not establish that any particular employer has approved redundancies.

The expected change also remains largely ahead of these companies. In the Deloitte CFO Survey, 67% of finance chiefs judged their organisations to be around a quarter of the way towards the transformation they ultimately expect from AI; Deloitte UK chief executive Darren Graves cautioned, “Capturing that potential will take more than adopting new technology.” The progress figure is the respondents’ assessment of their own organisations, not a measure of how many employees currently use AI each day.

What adopting businesses have experienced

The official business survey measures an outcome across a broader population. Its latest wave was voluntary and ran from 21 September to 4 October, with 9,933 responses from a sample of 38,592 businesses. That response total covers the whole survey wave, not just the businesses answering the AI headcount question. Estimates are weighted by business count and classed as official statistics in development.

The ONS questionnaire asks, “What impact, if any, have artificial intelligence technologies had on your business's overall workforce headcount?” Its response options include increased headcount, no change, reduced headcount, not sure and not applicable. This asks a business to assess the effect on its total workforce. It does not count individual dismissals or independently establish the reason each job ended.

AI use is more common than a reduction in staffing. In September, 30% of all surveyed businesses said they used at least one AI technology; among those with 10 or more employees, the share was 38%, rising to 51% for businesses with 100 or more employees. Among users in the group with at least 10 employees, 38% identified administrative or clerical roles as the most affected and 24% identified creative or design roles. A role being affected describes a change in work, which is distinct from a change in the number of people employed. The related finding on clerical roles and headcount explores that distinction.

Fewer than 1% of AI-using businesses in that size group said their headcount had increased as a result of adoption. The shares reporting increases and reductions had remained broadly stable over the preceding year. The figures are shares of businesses giving each answer, rather than shares of UK jobs created or lost.

Why the percentages measure different things

A comparison of the survey populations highlights the central distinction: major-company finance chiefs were asked about the coming year, while businesses already using AI were asked about an effect to date. The percentages therefore cannot be combined into a single trend or treated as competing estimates of the same outcome.

Business coverage matters too. The official survey is weighted across private sector businesses and excludes some industries, including finance and insurance; the CFO poll is concentrated among large employers. Neither percentage states what proportion of UK workers have lost their jobs because of AI. A small share of businesses could account for many employees, while a large share of businesses could each make a small staffing change.

The distance between expectation and outcome may narrow if adoption becomes more embedded and firms act on their plans. It could also persist if productivity gains alter tasks, vacancies or hiring plans without reducing total headcount. The surveys do not distinguish those future paths today.

The employment test ahead

Further business survey readings can show whether more AI users begin to report a reduction in their overall workforce. Changes in the most affected roles will matter alongside that measure: clerical work may shift before an employer’s total payroll does. For now, the concrete employment effect captured by the official survey remains limited, while the larger-company forecast concerns decisions still expected over the coming year.

Also read:

• AI Hits UK Clerical Roles First—Headcount Cuts Remain at 6%
• HubSpot Cuts 660 Jobs—but Says AI Did Not Directly Replace Them
• Bossware Is Already Scoring Workers—41% Say Nobody Told Them