Coding jobs for new graduates fell fastest last year. AI is the most likely reason.

The 2024 cohort of UK computer science graduates was the first large class to enter a software labour market where employers had working generative AI tools at scale. The coding-and-programmer placement rate for new graduates fell from about 40% to 28%, the sharpest drop of any graduate occupation in the latest HESA Graduate Outcomes Survey of more than 350,000 respondents surveyed 15 months after finishing their degrees.

The share of UK computer science graduates moving into any professional role dropped from more than 60% two years ago to 50% in the latest cohort, according to the 2027 Guardian University Guide, which uses the HESA data. Economics graduates saw employer demand for the well-paid targets, economists and management consultants, soften in the same window.

The "fastest-falling" framing is a ranking, not an absolute volume claim: more coders and economists are still being hired than several other graduate occupations. But it is the first time a UK graduate-outcomes dataset has caught a clear, named-field compression at this speed.

Matt Hiely-Rayner, director of the data consultancy Intelligent Metrix, which compiles the Guardian guide, attributes the shift to AI. "With AI providing such responsive and cheap grunt work in this field, it's hard not to conclude that this is behind the trend," he said.

Charlie Ball, head of labour market intelligence at Jisc, the UK education and data agency that co-runs the Graduate Outcomes Survey, is more cautious. "It is clear that something did happen to the software developer labour market last year," he said. "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."

That hedge matters. The dataset cannot cleanly separate AI from the post-2022 tech correction, the interest-rate cycle that thinned entry-level hiring across the economy, and the UK's post-Brexit visa restrictions on skilled-worker routes. Both named experts name AI as the most likely driver of the coding-specific drop, but the survey is not a clean causal instrument.

Computer science graduates are diversifying into cybersecurity and network engineering, both of which still require hands-on judgement and access controls that current AI tools handle less reliably. The shift is visible in the HESA activity breakdown for the 2024 cohort.

For students choosing courses this winter, the signal is that entry-level cognitive grunt work, the parts of a coding or analyst job that look like translation, summarisation, or first-draft writing, is the section of the labour market most exposed to current AI capability. Universities are now racing to fold more cybersecurity, systems, and applied-AI coursework into computer science degrees so graduates leave with skills further up the value chain. Employers, for their part, are rewriting entry-level job descriptions to ask for code review, model evaluation, and infrastructure work rather than greenfield feature shipping.

The next measurement point is the 2025 leaver cohort, surveyed in late 2026. If the coding-placement rate falls another ten points, the AI-causal reading will harden. If it stabilises, the post-2022 correction and rate cycle will look like the larger story, with AI one input among several.