From 40% to 28%: UK Data Shows AI Is Eating Computer Science Graduates' First Jobs

For years, the argument that AI would displace white-collar workers ran on anecdotes, projections, and CEO speculation. On September 12, the Guardian published something more concrete: national survey data showing that computer science — the degree that was supposed to be the safest bet of the AI era — is now the subject of one of the sharpest graduate employment declines in the UK.

The figures come from the data behind the 2027 Guardian University Guide, compiled by Matt Hiely-Rayner of Intelligent Metrix from the Higher Education Statistics Agency’s Graduate Outcomes survey, which gathered responses from more than 350,000 former students fifteen months after they finished their courses in 2024. The headline finding: coding and software development were the fastest-falling occupations for UK graduates last year.

The numbers behind the headline

Two data points carry the story. First, the proportion of computer science graduates finding professional jobs as coders or programmers has fallen from roughly 40% historically to just 28% in the latest survey round. Second, the proportion of computer science graduates moving into any graduate-level occupation at all has declined from more than 60% two years ago to 50% — meaning half of the UK’s CS cohort is now starting their careers below the graduate employment line.

“With AI providing such responsive and cheap grunt work in this field, it’s hard not to conclude that this is behind the trend,” Hiely-Rayner told the Guardian.

That is a striking statement from the person who compiles the tables. It’s one thing for a startup founder or a venture capitalist to claim AI is absorbing junior work; it’s another for the analyst behind one of the UK’s most-cited university rankings to look at a 350,000-person longitudinal dataset and point at AI as the most plausible cause.

The experts hedge — but only slightly

Charlie Ball, head of labour market intelligence at Jisc, the UK’s technology agency for education and research, offers the careful version: “It is clear that something did happen to the software developer labour market last year. While I’m reluctant to say that AI did it, because there’s little evidence, it’s very hard to escape the conclusion that AI has played a role.”

Ball notes that computer science graduates are diversifying into adjacent roles such as cybersecurity and network engineering — occupations that still require human judgment and where demand remains firm. “It’s not collapsed entirely, there are still roles available, but pure coding jobs seem to have been in particularly short supply in 2025,” he said.

That distinction — “pure coding” versus everything else — may be the most important sentence in the report. The jobs vanishing fastest are precisely the ones AI agents do best: well-specified implementation tasks, boilerplate, first-draft code. The jobs absorbing the displaced graduates are the ones wrapped in operational context, physical infrastructure, or adversarial reasoning.

Economics graduates are next in line

The data also shows a second, quieter signal: demand for economics graduates in well-paid financial roles — economists, management consultants — declined as well. Ball has been asking finance employers directly about their AI usage. “Most are saying they are not replacing jobs with it, they are changing jobs. They expect new hires to have familiarity with AI,” he said. “But it is a fast-moving field and employers are not unanimous. Some have an appetite for risk and are embracing AI tools and laying people off, in the hope that it will give them an edge.”

The pattern matches what analysis pieces at The Economist and others have flagged since spring: the entry rung of analytical professions is thinning simultaneously across coding and finance — the two domains where AI output is cheapest to substitute for a junior hire’s output.

Context: a graduate market already at decade lows

The University Guide data lands on top of an already weak market. In late August, jobs platform Adzuna reported that only 8,383 graduate vacancies were advertised in the UK in July — down 45% year-on-year and the lowest in a decade. Combine a shrinking vacancy pool with a reallocation of the remaining jobs away from junior technical work, and the class of 2024 is confronting a structurally different market than the one their course prospectus described.

There is a global parallel worth noting. Just days ago, reporting on China’s record 12.7 million graduates documented the same mechanism — AI absorbing the junior white-collar tasks that once trained new professionals — in a much larger labour market. The UK data now gives that narrative a hard statistical anchor in a Western economy.

Universities are starting to move

The Guardian’s 2027 tables place Oxford, Cambridge, and the LSE in the top three overall, with the University of Birmingham rising to 14th — its highest ever position. But the more strategically interesting move is curricular: Birmingham now lets undergraduates add an “intercalated year” in subjects such as AI and data science without the usual prerequisites.

“This is about equipping people for the world they are in, whether or not we like the world we are becoming,” said Birmingham vice-chancellor Adam Tickell. “Our young people need to understand how to use these things properly.”

What this means

One year of data is not a trend, and the survey measures the 2024 cohort during a 2025–26 market — a period of exceptional AI capability gains. But the direction is unmistakable, and it arrived where it was least expected: not in copywriting or customer support, but in the degree program that was explicitly sold as the AI-proof path. For students, the practical read is that “learn to code” is no longer sufficient — the durable entry-level roles are shifting toward security, infrastructure, and the judgment-heavy work of directing AI tools rather than competing with them.