The AI Pincer: Why Are British Firms Cutting Jobs While US Firms Hire?

British businesses are caught in a pincer right now, and the unemployment figures are finally starting to catch up.

One jaw of the pincer is the anaemic domestic economy. Firms are still absorbing the well-documented rise in employer National Insurance^[1], plus a National Living Wage that rose another 4.1 per cent to £12.71 in April^[2]. Now we have another energy shock to contend with. Wholesale gas prices rose 78 per cent between July and mid-September. The Bank of England expects inflation to top 4 per cent early next year, and three members of its Monetary Policy Committee voted last month for a rate rise^[3].

The official data shows reality is starting to set in. More than 100,000 fewer people are on UK payrolls than a year ago, and vacancies have dropped to 702,000, a level last seen outside the pandemic in 2014^[4]. Even these figures are contested, given the questions over how the data are collected, but the ONS’s own survey suggests smaller firms are holding off recruiting because of labour costs^[4]. That matches what we are hearing anecdotally in conversations across the tech sector.

The other jaw of the pincer is ever-expanding competition. Somewhere in your sector, a startup with just a handful of people, or a more forward-thinking competitor, is using AI to cut the cost of winning and serving customers. Quotes go out in minutes and invoices reconcile automatically. Support queues are cleared overnight by AI agents. This competitor can price lower, respond faster and take on more work without adding headcount. Standing still against them is a sure-fire way to go out of business.

Caught inside the pincer, British firms are making a predictable choice. In January, Morgan Stanley published a survey of companies that had used AI for at least a year. UK firms reported productivity gains of 11.5 per cent, in line with their American peers. The difference is in how those companies are using the gains. American firms, reaching for the stars, created more jobs than AI removed. British firms reported a net 8 per cent loss, twice the international average, with early-career roles the most likely to go^[5].

The technology itself is the same on both sides of the Atlantic, so why such a big difference in response to this paradigm shift? The economics around it differ, of course, but the divergence may come down to risk appetite.

US founders want to become unicorns, and they have access to capital to match their enthusiasm and ambition. The UK is no doubt a great place to start a company. Elite talent exists all over the country, and the administrative burden is very limited, at least compared with its European neighbours. But scaling is extremely hard, and access to capital is very limited. At the HumanX AI conference in Amsterdam recently, the majority of UK founders we spoke to had already made plans to move abroad, sell to a foreign investor, or both.

Graduates have already started to feel the pinch too. Adzuna listed 8,383 graduate vacancies in July, a record low, down almost half on a year earlier^[6]. Job-board counts may overstate or even understate the fall^[7], but Indeed also has graduate postings at their lowest for the time of year since 2020^[8]. Some of this is standard right-sizing, but even that is a response to these same market dynamics.

What should a squeezed business do?

First, find out where your people’s time is spent. In the firms we advise, a striking share of skilled staff time goes on repeatable process work: chasing, re-keying, reconciling, reporting. This is where the cost of sale hides, and AI agents handle it well.

Second, decide in advance what the savings are actually for. A business that ‘banks’ them has bought a one-off reprieve. A business that puts them into sharper pricing, faster delivery and more capacity has bought future growth. It is the latter that keeps hiring.

Third, keep the bottom rung of the ladder. Redesign junior roles around checking, correcting and supervising AI output. That work builds judgement faster than the data entry it replaces.

Finally, remove bloated middle management. Make sure senior managers embrace AI as a tool for leverage, pair them with a more dynamic junior base, and you will find that middle management adds little to the production line.

Government also has a part to play. An employer pays £3,750 a year in National Insurance on a £30,000 salary^[1], and nothing on the AI being deployed to replace that role. John Healey delivers his first Budget as Chancellor on 28 October^[9]. An incentive for firms that put AI savings into training and hiring young people is likely to cost less than the alternative: lost income tax and National Insurance, and a bigger welfare bill, for every entry-level job that never gets created.

The pincer keeps closing. Firms can use AI to shrink to fit a weak economy, or they can use it to grow out of one. Only the second choice will benefit the UK economy.