Teaching Career Guidance in the AI Era: What Changes for Teachers

This article is not written for parents or students. It’s addressed to education professionals (form tutors, psy-EN careers counsellors who are the French Education ministry’s guidance psychologists, teaching coordinators, independent guidance coaches) who now have to support students whose future job will look nothing like what it was five years ago.

I’m not a teacher. I’m an entrepreneur, and I code with AI every single day. That dual vantage point is what I’ll draw on here: what I actually see changing in the world of work, and what it means for the people who prepare students to enter that world.

The problem you’re facing

The careers-guidance role is probably the most affected by the changes underway, and yet it’s the one we talk about least when AI in education comes up. The spotlight is on “AI replacing maths teachers,” while the real, quiet challenge is this one:

How do you advise a 17-year-old on which pathway to choose, when the job they’re aiming for may look different at 23 and more different still at 30?

The classic guidance tools (ONISEP, France’s national careers information service, along with job factsheets and RIASEC tests) were designed for a world where jobs changed slowly. In 2010, a technology teacher or a psy-EN counsellor could say “banking careers are doing well” and be right for ten years. Today, the same statement has a shelf life of two to three years at most.

This doesn’t mean your job is disappearing. It means it’s changing in nature.

The three observable shifts

1. From the job to the capability

Before: “What job do you want to do?” After: “What capabilities do you want to build?”

The jobs that will be upended are the ones built on automatable tasks. The ones that will endure are built on irreplaceable human capabilities: judgement, negotiation, empathy, creativity, solving fuzzy problems.

What this means for your practice: instead of having the student pick a job, help them identify the capabilities they want to build and choose a course that builds them. A science prépa (France’s selective preparatory classes) builds the ability to reason under pressure. A history degree builds the ability to construct an argument. A work-study BTS (a two-year vocational higher diploma) builds the ability to solve concrete problems as a team. Whatever job they have in mind at 17, these capabilities will serve them in jobs that don’t exist yet.

2. From prediction to resilience

Before: “Which pathways have the best job prospects in ten years?” After: “Which pathways produce people able to adapt across three or four different jobs over a career?”

Career paths are fragmenting: in France, more than one worker in five changed occupation within five years (Dares, the French labour ministry’s statistics office). A 2026 graduate should expect several significant career changes over their working life. The question “which job” therefore matters less than the question “how much resilience.”

What this means for your practice: champion courses that keep several doors open rather than ones that close many to open a single one. A general bachelor’s degree followed by a specialised master’s is often more robust than a hyper-specialised BTS that locks a student into one specific job (even if it has an excellent employment rate within six months).

3. From the “career plan” to the “learning plan”

Before: “Build your career plan.” After: “Build your learning plan.”

A career plan at 17 is an exercise in projecting into the unknown, one that often breeds anxiety without producing a good decision. A learning plan, by contrast, is a commitment for the next three to five years that is both more concrete and more realistic.

What this means for your practice: ask less “what do you want to do later?” and more “what do you want to learn over the next three years?” The second question yields better answers because it’s actionable.

How to use AI as a tool (not a threat)

Generative AI is an excellent tool for preparing a guidance meeting, provided you know how to use it. Here are three concrete uses I recommend to educators.

Use 1: explore a job in ten minutes

Before a meeting with a student considering a job you don’t know well (data scientist, UX researcher, product manager, prompt engineer, and so on), ask the AI:

“Describe the job of [X] in 2026: day-to-day tasks, required skills, typical path into it, average salaries, five-year outlook. Cite your sources.”

In 30 seconds you get a briefing that would have taken 30 minutes to piece together on Google. The AI makes mistakes, but so do your students, and at least you walk in with a structured basis for discussion.

Use 2: simulate a pathway

Give the AI a student’s profile (age, chosen specialist subjects, grades, interests) and ask it:

“Suggest three different post-Bac pathways for this student, and for each one: course, duration, cost, typical outcomes, risks.”

The suggestions aren’t perfect, but they’re often more varied than the ones a teacher produces off the cuff from their usual assumptions. It opens up the conversation with the student.

Use 3: unpack a parental narrative

A parent tells you “my child has to study medicine, it’s stable and well paid.” Rather than sparring bare-handed with statistics you don’t have to hand, ask the AI to produce a fact-based counter-argument on the real state of medical studies in 2026 (failure rates in PASS, France’s first-year medical access pathway, workload, working conditions after residency, actual early-career salaries). You use it as material for the conversation with the parent, not to prove them wrong, but to add nuance.

What AI doesn’t do (and won’t)

AI cannot:

• Know your student. It doesn’t see their body language, their stress, their silences, their unspoken worries. You do.
• Manage family dynamics. A three-way guidance meeting (student plus two parents who disagree) is a human negotiation that AI cannot arbitrate.
• Take responsibility for advice. When you recommend a path to a student, you put your professional judgement on the line. AI does not.
• Support someone over time. A guidance relationship is built over several meetings. AI has no continuous memory of the student.

It’s in these four dimensions that your professional value is irreplaceable. Not in producing job factsheets (AI does that better and faster), but in the human quality of the support you provide.

Three practical recommendations

1. Learn to use ChatGPT, Claude, or Gemini at your own comfort level

You don’t need to be a developer. Open a free account, ask it questions about the jobs your students are considering, test it on your own subject areas. In two hours you’ll have a clear sense of what it does well and what it does badly. That’s the only way to know when you can trust it and when you can’t.

2. Keep a deliberately non-productive human margin

AI will speed up your research and writing tasks. Use the time you save to spend more time with each student, not to see more of them. That’s the only way to preserve what makes you valuable.

3. Talk about it openly with students

Teenagers in 2026 already know how to use ChatGPT for their homework and to think through their options. If you act as though it doesn’t exist, you come across as out of touch. If you talk about it openly, you show them you take the tools they use seriously, and you can teach them to use those tools better.

The message I want to get across

The careers-guidance role is probably more important than it has ever been, not less. In a world where the labour market is transforming fast, students need, more than ever, an experienced human adult to help them find their way, not a static job factsheet or an algorithm.

But the role has to evolve. Less “which job” and more “which capability.” Less prediction and more resilience. Less of a fixed career plan and more of a living learning plan. And AI is a tool that can help you do this work better, not a threat that replaces you.

Key takeaways

• Your job is not threatened by AI: it’s being transformed by it, like every job that rests on human judgement.
• Three shifts to absorb: from the job to the capability, from prediction to resilience, from the career plan to the learning plan.
• AI is an excellent preparation tool for your meetings, not a replacement.
• What AI won’t do: know your student, manage family dynamics, take on professional responsibility, support someone over time.
• Your irreplaceable value lies in the human quality of the support, not in producing job factsheets.
• Talk about AI tools openly with your students: they’re already using them, and you can teach them to use them better.

Going further

• Will AI really replace the jobs my child is preparing for today?
• Choosing a career in 2026 when you don’t know what jobs will look like in 2040
• OECD: Employment Outlook
• World Economic Forum: Future of Jobs Report

---

Article written by Constantin Mardoukhaev, co-founder of Axiom Academic. Constantin codes with AI every day and supports French-speaking families with their guidance decisions, including through regular conversations with educators looking to adapt their practice.