AI is reshaping graduate roles, not eliminating them, ISE survey finds

TL;DR:

• An Institute of Student Employers (ISE) Student Development Survey finds 87% of UK employers expect AI to change entry-level roles, but four in ten expect no graduate jobs to be replaced over the next three years; a similar proportion expect only a small number of roles to be affected.
• Two-thirds of employers are concerned that candidates are using AI in ways that misrepresent their abilities during selection; over half have increased the time devoted to professional-skills induction.
• Resultsense view: this is a useful counterweight to the dominant “AI ate the grad job” narrative — and lands the same week Lord O’Donnell argued AI tax revenue should fund mass retraining and YouGov polling showed 47% of Britons would back an AI tax. Reshaping rather than replacing has different fiscal and policy implications than displacement.

The Wonkhe analysis of the survey, written from the higher-education perspective, frames the story around an analogy to the 1980s Fleet Street printing dispute. The argument is familiar — historical technology shifts have tended to raise the skill content of work rather than eliminate it — but the current data is fresher than the analogy suggests.

What the survey shows

Most employers anticipate adjustments to tasks rather than wholesale headcount cuts. Routine activities — basic research, drafting, administrative processing — are becoming less central to entry-level roles. AI capability is shifting to a baseline expectation rather than a specialist skill, with a majority of employers building AI literacy and practical use into training programmes.

The skills profile is rotating. Employers are placing increasing value on critical thinking, judgement, adaptability and communication — capabilities that let graduates interpret and apply AI outputs rather than simply produce them. While employer satisfaction with foundational graduate skills remains broadly steady, one-third rate graduates as below expectations on entry in adaptability, self-awareness, and awareness of wider organisational context.

The recruitment integrity problem

The headline-grabbing finding from a hiring perspective is that two-thirds of employers are concerned candidates are using AI in ways that misrepresent their abilities during selection. The response inside firms has been visible: over half have lengthened professional-skills induction time to bridge the work-readiness gap.

UK higher-education implications

Wonkhe argues universities should embed AI literacy across the curriculum rather than treating it as a specialist add-on, with greater emphasis on judgement, adaptability and self-awareness. More than 90% of employers in the survey believe placements contribute significantly to student development — reinforcing the case for embedding work-based learning across the student experience.

Cross-source context

The survey lands against several parallel UK signals this week. Master of the Rolls Sir Geoffrey Vos urged lawyers and judges to “embrace AI to stay relevant” to the TikTok generation. Liz Kendall, the science secretary, set out a Sovereign AI venture fund and a forthcoming AI hardware plan. And the SolarWinds public-sector IT survey published the same day shows 56% of public-sector IT roles have become more demanding because of AI — a reshaping pattern, not a displacement one.

Looking forward

For UK universities and employers, the operational message is that the entry-level pipeline can adapt — but only if the curriculum and induction processes update at the same pace as task structures change inside firms. The political message, meanwhile, is that the case for an AI tax to fund retraining sits uneasily alongside data showing reshape rather than replace. Both narratives can hold; whichever Whitehall picks will shape skills policy through the rest of the decade.