Is a Degree Still the Entry Ticket? What Employers Now Look for in Graduates.

IN ONE MINUTE

A degree still opens doors, but it is no longer enough on its own to prove readiness. AI is absorbing much of the routine work graduates used to learn on, and employers are raising the bar on judgement and communication. In Monster's 2026 research, 89% of graduates worried AI could replace entry-level jobs, up from 64% a year earlier, and only 36% felt college was preparing them to use AI at work. Graduates stand out by showing evidence of judgement: real projects, work they can defend, and the ability to spot when AI is wrong.

THE OBJECTIVE

Several university professors have asked me what they should tell students about today's job market. This issue is my honest answer, for students, parents, universities and employers.

WHAT THE RESEARCH SHOWS

• Monster (2026): 89% of graduates worry AI could replace entry-level jobs (64% in 2025). Half have used AI in coursework, but only 36% say college is preparing them to use it at work.
• ResumeTemplates.com (Jul 2026, 1,000 US hiring managers): 48% said they would rather invest in AI tools than hire and train a recent graduate.
• Indeed via the FT (Aug 2026): UK job postings open to new graduates fell about 7% year on year, the lowest for the time of year since 2020.

Korn Ferry's David Farris describes what is emerging as a "fat pyramid": with AI handling basics like scheduling and data entry, early-career professionals can step into more strategic, high-impact work sooner.

That can be good news for capable graduates. It means earlier responsibility. But the old path of "get the degree, get any junior job, learn quietly for two years" no longer describes the market.

THE FRAMEWORK: THREE PROOFS OF READINESS

Proof	What it shows	How to build it
1. Applied work	You can deliver in the real world	Internships, live projects, independent work you can explain and defend
2. AI judgement	You can use AI and catch it when it's wrong	Show how you checked, corrected or challenged an AI output
3. Professional basics	You can communicate and handle real documents	Clear writing, reading a budget or contract, presenting to a senior audience

WHAT THIS MEANS FOR STUDENTS AND RECENT GRADUATES

If I were advising a graduate directly, here is what I'd tell them.

Treat AI fluency as a baseline, not a differentiator. Half of graduates have already used AI in their coursework, so using it is quickly becoming table stakes. The differentiator is judgement: can you tell when the AI's output is wrong, incomplete or subtly misleading? Employers are not looking for people who can prompt a model. They are looking for people who can catch it when it is confidently incorrect.

Build a portfolio of demonstrated judgement, not just credentials. A transcript shows what you studied. An internship, a real project with a messy real-world outcome, or a piece of independent work you can defend under questioning shows how you think. Given the choice between a slightly stronger grade and a real, if modest, body of applied work, build the applied work.

Practise the professional basics deliberately. In a ResumeTemplates.com survey, nearly 70% of hiring managers reported at least one character concern about recent graduates. 41% said graduates struggled to write a professional email, and 76% said they needed help with documents such as memos, contracts and budgets. These are basics, and employers are now testing for them explicitly.

Seek roles that give you real responsibility early, even at a smaller company. If entry roles now ask for strategic thinking sooner, look for employers who are open about that and set up to support it, rather than ones still running a two-year ramp-up on a role that no longer has two years of ramp-up work in it.

Don't mistake AI confidence for AI competence. Being comfortable using AI tools casually is not the same as directing them at a genuinely difficult problem, or knowing when to distrust the answer. In Korn Ferry's 2026 Talent Acquisition Trends survey, 73% of talent leaders named critical thinking their No. 1 priority when evaluating hires.

WHAT THIS MEANS FOR UNIVERSITIES

To the professors who asked me the original question: I believe the role of the university is not shrinking. It is being asked to do something harder.

For most of the last century, universities were the main route to structured knowledge. That has changed. A motivated student now has a patient, always-available tutor for almost any subject, at almost any hour. That is not a threat to resist. Used well, it should raise the floor of what every student can learn.

What AI cannot yet do, and where I think the university's real value now sits, is teach a young person to stay with an unresolved, ambiguous problem long enough to produce something new. Original research, contested argument, and the experience of being wrong in front of peers and revising your thinking in real time are not content-delivery problems that AI solves. They are about formation, and they still need human mentorship, real deadlines and the energy of a cohort.

The universities that make this shift will spend less time as the main source of information and more time as the place where students apply it under pressure, defend it under challenge, and build the judgement no model can hand them. Their graduates will enter the market with exactly the skill employers say they need most: independent judgement under ambiguity.

WHAT THIS MEANS FOR EMPLOYERS AND HR LEADERS

The honest counterpart: many traditional hiring filters are becoming weaker predictors of the thing they are meant to predict.

A degree from a well-regarded university has long been used as a proxy for ability and diligence, but it was never a direct measure of judgement. It is an even weaker proxy now that many of the hardest parts of junior roles are being automated. A polished CV and a confident interview are also, in part, a test of how well a candidate used AI to prepare, which tells you less than it used to about how they will perform.

If your organisation still screens mainly on grades, university reputation and interview polish, it is relying on signals that are losing predictive power, just as the roles themselves demand more judgement, sooner. The answer isn't to abandon structured hiring. It is to redesign what you test for.

In practice, that means:

• realistic work-sample assessments over credential screening
• explicitly testing a candidate's ability to critique and correct AI output, not just produce work with AI's help
• treating the first year of employment as a capability investment, not a ramp-up cost

Organisations cutting entry-level hiring to save cost today may be weakening the leadership pipeline they will need in ten years.

THE LAST WORD

The degree still gets you in the room. It no longer gets you the job. What gets you the job is proof you can think clearly when the AI can't, or won't.

If you're hiring graduates this year, or sending one into the market, what's the one skill you wish universities spent more time teaching?

Next week: the sequel question. If AI has changed how graduates get hired, what happens to the people already inside your organisation, and who is training them now?

—

Until next week,

Artificial Ian

Tell your network: if you know a graduate, a parent or a professor thinking about this, this one is for them.

→ Found this useful? Forward it to a professor, a hiring manager or a graduate.

→ Subscribe to The AI Executive Briefing, weekly in your LinkedIn inbox.

Sources: Monster Graduate AI Readiness Report (2026); ResumeTemplates.com hiring manager survey (Jul 2026); Indeed Mid-Year UK Labour Market Update, via the Financial Times (Aug 2026); Korn Ferry, Talent Acquisition Trends 2026 (Oct 2025); Korn Ferry Institute, "Early Careerist: The AI On-Ramp" (Oct 2025).

Credit: David Farris and Emily Gianunzio, Korn Ferry Institute.