OddsFrontUK data shows a sharp decline in graduate coding jobs as AI reshapes recruitment
Computer-science graduates are entering programming roles at a markedly lower rate, although official analysis cautions that artificial intelligence has not been proved to be the sole cause.
Graduate outcomes deteriorate
New UK graduate-outcomes data show a pronounced decline in the proportion of computer-science graduates moving directly into programming and software-development work. About 28% of recent graduates entered those occupations, down from roughly 40% in the previous comparable cohort. The wider share obtaining any graduate-level position also fell to about half, from more than 60% two years earlier.
A large cohort behind the figures
The findings draw on the Higher Education Statistics Agency's Graduate Outcomes survey, which followed more than 350,000 former students about 15 months after they completed courses in 2024. The figures were analysed for the 2027 Guardian University Guide. Software and coding emerged as the fastest-falling graduate occupation, making the shift more substantial than a routine annual fluctuation in a single employer or region.
AI is changing the entry point
Recruiters increasingly expect applicants to understand how generative systems can accelerate coding, testing and documentation. At the same time, some tasks historically assigned to junior developers can now be performed or assisted by those systems. That combination may be reducing traditional entry-level openings while raising the technical and evaluative skills demanded from the graduates who do secure them.
Causation remains unsettled
Neither the university data nor the government's labour-market assessment establishes that AI alone caused the decline. Hiring weakened more broadly, economic conditions changed, and employers adjusted staffing after an earlier technology expansion. Official analysis nevertheless identifies computer programming as a particularly exposed field and reports a steep fall in employment among younger workers in that occupation.
The policy problem
The result creates a mismatch between expanding university interest in computing and a narrowing route into professional experience. If employers remove too many junior positions, they may also weaken the pipeline through which future senior engineers acquire practical judgment. Universities face pressure to teach system evaluation, domain knowledge and collaborative problem-solving alongside conventional programming rather than treating AI familiarity as a separate optional module.
What to watch next
The next Graduate Outcomes release will show whether this is a temporary hiring correction or a structural change in how software teams are built. Other indicators include the volume of entry-level advertisements, salary differences between AI-related and conventional roles, and whether employers create new junior jobs focused on checking automated output. Any policy response will need to distinguish displacement caused by technology from the effects of the wider economic cycle.