Can picking the right degree protect your job from AI?

A UK expert on graduate employment says futureproofing a 45-year career is 'really quite hard', but face-to-face work still looks safer than most.

AI2Day NewsdeskEditor: Lee Brown3 min read
Photoreal news-editorial 16:9 image of a large federal government building exterior at dusk, dramatic low-angle shot showing imposing stone columns and windows
Share

Key points

  • Charlie Ball, a graduate employment expert at Jisc (the UK's higher education digital, data and technology agency), says futureproofing a full career against AI disruption is "really quite hard."
  • Ball identifies jobs built around face-to-face human interaction as the category least likely to be replaced by AI.
  • No single degree is a guaranteed shield; the pace of technological change makes long-range predictions unreliable.

AI systems that can do white-collar work, sometimes called AI agents (software that carries out multi-step tasks without a human in the loop), are arriving faster than most university prospectuses have caught up with. Students choosing courses today are betting on what the job market will look like in the 2060s.

That's a bad bet to have to make.

"What you're trying to do is futureproof a 45-year career in a time of rapid technological change," Charlie Ball told The Guardian, "and that's really quite hard to do."

Ball works on graduate employment for Jisc, the body that supports digital, data and technology services for UK universities and colleges. His view carries weight because he tracks what happens to graduates once they leave campus. We've been following the same question since July, most recently in our 7 September story on why designers are holding their ground against AI.

So which jobs hold up?

Ball's short answer: roles where the point is the human being in the room. Nurses taking a patient's hand. Teachers reading a classroom. Social workers making a judgement call in someone's living room.

His reasoning is direct. AI handles information well. It processes, summarises and codes. What it cannot do, at least not yet, is physically show up, build genuine trust over time, or carry the legal and moral weight of a licensed professional making a high-stakes call about another person.

That's not a permanent firewall. It is a meaningful buffer.

What does this mean for students right now?

Choose a field where human contact is the product, not just a feature. A degree in nursing, social work or counselling puts you in a category where human presence is the whole point. No course is immune to change, though. Even heavily human roles will use AI tools within the next decade, shifting what practitioners do day-to-day. A nurse who understands how an AI diagnostic tool, software that flags early warning signs in patient data, reaches its conclusions will be more useful than one who doesn't.

Ball's point is modest but real: pick work you'd want to do anyway, in a field that requires being present with people, and plan to keep learning throughout your career rather than treating graduation as the finish line.

Common questions

Is any degree completely safe from AI disruption?

No expert interviewed by The Guardian or cited by Jisc has claimed any field is untouchable. Ball's point is about relative risk: human-facing roles face a lower near-term threat, not zero threat.

Should I avoid degrees in tech because AI will replace those workers too?

Not necessarily. People who understand AI systems and can direct or audit them are likely to stay in demand. The risk is greater for roles focused on a single, repeatable task that software can learn to do faster.

© 2026 AI2Day