Preparing for 2040: Four AI Scenarios That Put Your Future in Your Hands

Here is a number that should change how you think about AI and your career: 27 percentage points. That is the gap between the 42% of European workers who say they need AI skills and the 15% who have actually received any AI training [Fact]. The future of your job is not being decided by robots. It is being decided by whether that gap gets closed — and by whom.

That is the quietly radical message buried inside a new foresight report from Cedefop, the European Union's agency for vocational education and training. Published in December 2025, "Preparing for 2040: four AI-powered scenarios for the future of continuing skills development" does something most AI coverage refuses to do. Instead of predicting one inevitable future, it lays out four — and then makes the case that which one arrives is a choice, not a prophecy.

AI is the wild card, not the verdict

Cedefop's researchers looked across every force shaping how adults will learn and re-skill over the next fifteen years, and one factor rose above the rest. "AI stands out as the most influential and yet unpredictable force shaping CSD by 2040," the report concludes [Fact]. Continuing skills development — CSD — is the unglamorous engine of every working life: the training, retraining, and guidance that lets a 45-year-old logistics coordinator become something else when the old job changes.

Notice the two words doing the heavy lifting: influential and unpredictable. Influential means AI matters enormously. Unpredictable means nobody, including the people building it, knows exactly where it lands. That combination is not a reason for dread. It is the entire reason human decisions still count.

Four futures, and only one of them is dystopian

The report sketches four scenarios for 2040, and it is worth sitting with each rather than skipping to the scary one.

Scenario A, "A future of opportunities," is the optimistic path: AI enhances human potential, drives inclusive growth, and creates purpose-driven careers [Fact]. Scenario B, "Left alone to ride the tide," describes a society split by AI shockwaves, where the responsibility for keeping up shifts almost entirely onto individuals and skills-based inequality hardens [Fact]. Scenario C, "Staying afloat," is the quiet disappointment — AI is adopted so slowly that major disruption is avoided, but most of its economic and social upside is missed too [Fact]. And Scenario D, "AI unleashed," is the dystopia: large-scale job destruction, AI deployed mainly for control and profit, worker rights eroded [Fact].

Here is the part worth underlining. Three of the four scenarios are not defined by how powerful the technology gets. They are defined by how societies choose to govern it and share it. Scenario D is not what happens when AI gets too good. It is what happens when people with power decide the gains belong only to them. That is a political and institutional outcome, not a technological one [Claim].

The gap that decides everything

This is where that 27-point number comes back. Cedefop's 2024 European AI skills survey — the first of its kind — found that 42% of workers believe they need to develop AI-related skills, yet only 15% have taken part in any AI training [Fact]. Around one in three European workers already uses or interacts with AI tools at work, and 61% agree they will need genuinely new knowledge and skills to cope with AI over the next five years [Fact].

Read those numbers together and the four scenarios stop feeling abstract. The training gap is widest for low- and middle-skilled workers, older workers, and women [Fact]. If that gap keeps widening, Scenario B — polarisation, everyone left to fend for themselves — becomes the default, not through malice but through neglect. Close it, and Scenario A moves within reach. The distance between the hopeful future and the divided one is, quite literally, measured in who gets access to learning.

The "red flags" are your levers

Cedefop does not end on the scenarios. It flags cross-cutting "red flags" that policymakers must confront so the technology delivers inclusive outcomes: AI's dual potential for empowerment versus power concentration, the transformation of workplaces, the reinvention of learning and career guidance, the individual-responsibility gap, a shifting employment landscape, and widening inequality [Fact].

Reframe those red flags and they turn into a to-do list. Each one names a place where a decision has to be made — by a government, an employer, a union, a training provider, or you. The "individual responsibility gap" is not a scold telling you to fix yourself alone. It is the report warning that leaving people to self-rescue is exactly how the bad scenarios win. The levers exist. The question is only whether they get pulled.

What this actually means for your job

If your work touches training, learning, or people development, this report is describing your next decade directly. Training and development specialists, whose entire craft is helping others adapt, sit at the center of every optimistic scenario — see the detailed outlook on training-development-specialists. The same is true for instructional-designers and education-counselors, whose guidance work Cedefop explicitly names as an area AI will reshape rather than simply erase.

Even outside the learning professions, the pattern holds. Human resources managers and workforce-planning-managers are the people who will decide, inside thousands of ordinary organisations, whether that 27-point training gap shrinks or grows. Their choices are the scenario mechanism in miniature.

And if none of those is your job? The message is the same, just more personal. The report's own logic says the single most protective thing an individual can do is refuse to be left "alone to ride the tide" — by demanding access to training, joining collective efforts that spread the gains, and treating AI skills as something you are owed support to build, not a private burden.

Hope is a strategy, not a mood

It would be easy to read a report with a scenario literally called "AI unleashed" and conclude the worst is coming. But that is a misreading. Cedefop went to the trouble of describing four futures precisely because it does not believe one is fixed. The pessimistic scenarios are warnings, not forecasts — and warnings only matter to people who still have time to act.

The honest caveat: this is a foresight exercise, not a measurement of what will happen, and scenarios are deliberately simplified sketches rather than predictions [Estimate]. But that is their strength. They hand you a map of the forks in the road while you can still choose a turn. The data on the training gap is the compass showing which way we are currently drifting — and it is drifting toward division unless someone changes course.

That someone includes you. Not as a slogan, but as the report's actual finding: 2040 is not a place we are being sent. It is a place we are building, one training decision at a time.

Sources

• Cedefop, "Preparing for 2040: four AI-powered scenarios for the future of continuing skills development" (December 2025)
• Cedefop publication landing page: "Preparing for 2040"
• Cedefop, "AI skills in the workplace? Cedefop asked the workers" — European AI skills survey, 2024

Update History

• 2026-07-22 — Initial publication, based on Cedefop's December 2025 "Preparing for 2040" foresight report and its 2024 European AI skills survey.

---

AI-assisted analysis. This article was drafted with AI support and reviewed by a human editor before publication. Figures are drawn from Cedefop's published foresight report and skills survey; scenario descriptions are quoted from the source. Statistics are labelled [Fact] where drawn directly from Cedefop data, [Claim] for our interpretation, and [Estimate] where they reflect projections or inherently uncertain foresight.