AI in Careers Services: Key takeaways from our Graduate Futures Institute panel discussion

Across the panel, the consensus was clear: AI is a practical toolkit, not something to fear. But it's no miracle solution either. Several panelists shared how AI has helped with their own work, from writing proposals to generating ideas, while being equally clear that outputs always need sense-checking. The accuracy issues are real.

For careers leaders, that means modelling critical evaluation for your teams. If you're using AI, show your team how you review and refine what it produces, highlighting the need for human input.

Many careers teams are seeing:

• CVs that look almost identical because they've all been generated by AI
• Tools designed to help candidates stand out are making everyone sound the same
• Applications claiming skills students haven't actually developed or practised

But it goes deeper than indistinguishable applications. AI magnifies what's already there. Students with genuine experiences and strong self-awareness can use AI to articulate themselves more effectively. Students without that foundation are more likely to produce generic content that exposes their gaps rather than filling them.

The takeaway? The reflective work careers services do with students, like helping them identify experiences, articulate strengths and build self-awareness, becomes more important in an AI world, not less. AI can help students communicate, but only if they have something genuine to communicate in the first place.

One example from the panel made this point clearly. During an assessment centre, a student photographed the brief, fed it into AI and announced he was done. When challenged, he argued he was demonstrating efficiency and problem-solving.

What followed was an unplanned but valuable conversation. Had he checked the employer's AI policy? Was he even allowed his phone in the room? What skills was he actually demonstrating to assessors?

The panel was clear: some employers actively embrace AI, but they're evaluating how candidates engage with it thoughtfully, not just whether they can use it as a shortcut. Technical AI literacy matters, but so does judgment about when and how to use it.

One of the most practical parts of the discussion focused on getting buy-in across a careers service. One institution has established an AI working group with representatives from each area. The key was starting not with "where can we use AI?" but with "what problems are we trying to solve?"

The group includes volunteers with genuine interest, and feeds into a university-wide AI forum for alignment across the institution. This problem-first approach avoids implementing AI for its own sake and keeps the focus on tangible service improvements.

If you're thinking about a similar approach, consider starting small: identify one or two pain points in your service, bring together interested colleagues, and explore whether AI can help. Build from there.

The panel explored how AI can support advisers directly. One example involved using AI during appointments with students from unfamiliar course areas. The adviser used it to quickly surface relevant sector information, then used that as a springboard for reflective questions.

Rather than pretending to know everything, the adviser and student explored together. AI filled the surface-level knowledge gaps, freeing the conversation to focus on deeper career thinking.

This reframes AI as something that enhances the guidance relationship rather than replacing it. The human skill is knowing which questions to ask next.

Another practical example: embedding AI reflection into existing career assessments. Students run their CV through an AI tool, then critically evaluate the output as part of the task.

This does two things. First, it normalises AI use so students don't feel awkward about it. Second, it develops the critical thinking skills they need to use AI well. Without that critical lens, students risk submitting content with inaccuracies, generic phrasing or claims they can't back up in an interview.

It also creates a natural opening for honest conversations about AI. When it's built into the process, the awkwardness disappears.

Students don't distinguish between AI use in academic work and AI use in careers applications. If the library says one thing, the business school says another, and the careers service says something else, students are left confused.

Careers services can't set AI policy in isolation. Effective implementation requires collaboration with academic departments, central university teams and student-facing services. If your institution doesn't yet have a joined-up approach, it's worth raising the question.

AI isn't replacing the human elements of careers guidance. The reflective conversations, the challenging questions, the ability to read context and respond with nuance – these become more valuable when AI can handle the routine.

For careers leaders, the question isn't whether to engage with AI. It's how to do so in ways that strengthen what makes careers guidance effective in the first place.

Thank you to our panellists: Noeleen Hammond Jones, deputy head of LUMS Careers at Lancaster University; Joel O'Toole, careers consultant at Birmingham City University; James Goodwin, head of careers at Aston University; and Bonnie van den Bergh, lead product manager at Access Careers Centre.