Middle England has been told, day after day, that the great menace to living standards is the influx of illegal migrants: pressure on wages, pressure on schools, pressure on the NHS, pressure on housing, and a gnawing fear that the ladder of opportunity is being pulled up for their children.
It’s a powerful story because it has a villain you can picture: a small boat, a hotel bill, a queue at A&E. But while the country argues about the faces we can see, something far bigger is slipping quietly through the front door, and it isn’t human.
Not the sci-fi kind with red eyes and killer robots. The dull, spreadsheet-and-email kind. The “write me a report”, “summarise these documents”, “draft the legal note”, “produce the slide deck”, “screen these CVs”, “generate the code”, “do the first pass analysis” kind. The kind of work that has, for decades, been the bridge between a good degree and a good life.
And here’s the uncomfortable truth: for the median British family, the office worker, the junior professional, the graduate parent watching their children line up for internships, immigration is not the main force squeezing pay, prospects and security. AI is.
What immigration actually does to middle-class living standards
Start with the thing most people assume, that immigration massively undercuts wages. The evidence, across decades of UK research, simply doesn’t support a big, broad wage hit. The Oxford Migration Observatory’s briefing on labour-market effects finds that average impacts on wages and employment are small, and where there are effects, they’re concentrated in particular groups and places rather than across the board.
Then there’s the fiscal claim, “migrants drain the system”. Again, it’s complicated, but the scale is often exaggerated in popular debate. The Migration Observatory’s review of the fiscal impact notes that studies typically find the overall fiscal effects of immigration are less than 1% of GDP.
Even on the numbers, the picture is shifting fast. The Office for National Statistics estimates net migration fell sharply: provisional long-term net migration for the year ending June 2025 was 204,000, around two-thirds lower than the year ending June 2024, which stood at 649,000.
Yes, illegal migration is real and politically combustible, and government data reported a rise in detected illegal arrivals in 2025 even as visa issuance fell. But if you’re a middle-income household worried about the next pay rise, the next promotion, or whether your child can land a decent graduate job, the research says immigration is not the giant invisible hand grinding you down.
So where is the pressure coming from?
The silent arrival that doesn’t need a visa
AI doesn’t argue. It doesn’t sleep. It doesn’t ask for a pension. It doesn’t join a union. It scales instantly. And crucially, it targets the very tasks that used to be the training ground for graduates.
Look at what is happening right now to entry-level and graduate hiring:
- The Institute of Student Employers reports graduate hiring fell 8% year-on-year, and 42% of employers reduced graduate hiring.
- High Fliers’ research says the UK’s top employers have cut graduate recruitment by almost a quarter over three years, and it’s forecast to reduce further.
- Job listings data reported in February 2026 shows graduate job ads falling below 10,000 for the first time since tracking began in 2016.
That is not a marginal change. That is a trapdoor opening under the next generation.
Now ask yourself, what work is vanishing? Exactly the kind of “junior” tasks AI is already doing passably well: first drafts, basic research, initial analysis, routine coding, document review, templated communication, compliance checks, slide building, and the endless administrative glue that once justified hiring bright young people to learn the ropes.
Even the official and international evidence is blunt about where AI hits hardest: white-collar occupations, typically requiring tertiary education, are among the most exposed.
And adoption is already embedded. ONS analysis indicates AI adoption is higher in services than other sectors, precisely where so many middle-class jobs sit. The ISE has highlighted just how fast generative AI has arrived in workplaces, with a significant share of workers already using AI tools by mid-2024.
This is the part politicians tiptoe around, because it scrambles the old culture-war script. You can restrict visas. You can patrol borders. You can deport people. But you can’t handcuff a software update.
Why this matters more than border headlines
Immigration’s effects, when they bite, show up in specific local pressures: housing demand in hot markets, stretched services where funding lags population, and competition in particular job segments. Those are real issues, and they deserve grown-up policy.
But AI is different: it’s a system-wide shift in bargaining power.
When a firm can replace ten junior analysts with one senior manager plus AI tools, the whole career ladder changes:
- fewer entry points
- slower progression
- more intense competition for the remaining roles
- and a brutal squeeze on wages at the bottom of the professional pyramid
That is how the middle class gets hollowed out, not by a single dramatic event, but by the quiet disappearance of the “starter” jobs that turn education into earnings, and earnings into mortgages, families, stability and a decent retirement.
If you want a single image, stop picturing a boat. Start picturing an inbox. The real pressure on Middle England isn’t coming from the Channel, it’s coming from the software quietly rewriting the rules of work.
What Middle England should demand, now
The answer isn’t to smash the machines or pretend AI will politely “create new jobs” at exactly the same rate and quality as the old ones. The answer is to fight for a fair transition:
- A national plan for entry-level work, incentives for firms to keep genuine junior roles, apprenticeships and paid training pathways, not “AI-assisted” excuses to cut headcount.
- Skills that match the new ladder. If the bottom rungs are disappearing, schools and universities must teach students to do what AI can’t easily do: judgment, client work, problem framing, accountability, and real-world execution, and employers must actually train them, many still don’t.
- Transparency from big employers. How many graduate roles are being cut because of AI, and where? If firms can quantify efficiency gains, they can quantify social costs.
- A tax-and-invest bargain. If AI boosts productivity while shrinking payrolls, the public should share in the upside, funding education, reskilling, and the public services that keep communities stable.
Middle England is right to worry about living standards. But it’s staring at the wrong suspect.
The border argument will rage on, because it’s emotional, visible, and politically useful. Meanwhile, the silent newcomer keeps walking into the workplace, taking tasks, taking rungs off the ladder, and taking away the one promise Britain has always made its strivers:
Work hard, get qualified, and your children will do better.
AI doesn’t need to cross a border to break that promise. It’s already inside.
Leave a Comment