16 November 2026
Candidate

AI and jobs in Switzerland: which roles are at risk, which are hiring, and how to position yourself

A developer wearing orange headphones works standing at two screens of code, in an office with large windows.

Since ChatGPT arrived in late 2022, unemployment has risen up to 27% more in the occupations most exposed to AI than in little-exposed ones, and job postings there have fallen by 38% in relative terms (KOF, ETH Zurich, October 2025). Among software developers, the number of unemployed almost tripled between late 2022 and late 2025. Over the same period, vacancies in skilled trades and technical occupations grew by 131% in ten years, against 30% for the market as a whole (Adecco and University of Zurich, 15 October 2026).

This guide does not predict the end of work. It brings together what Swiss figures actually show in 2026: which occupations are exposed, which are hiring, what is changing for newcomers, and how to position your job search when AI is part of every role. Every figure is dated and sourced, because this is the topic where approximations are most common.

Just starting out? Our guide to a first job in Switzerland gives the figures on entering work after studies or an apprenticeship, what recruiters expect and your unemployment rights.

AI and jobs in Switzerland in figures

The measures do not all say the same thing, because they do not measure the same thing: tasks, jobs, postings or unemployed people. Here they are side by side:

What is measuredResultSource
Use of generative AI in the population43% of 15 to 88 year olds used it in spring 2025, 31% for professional purposes; 79% of 15-24 year oldsFSO, 5 December 2025
Unemployment in highly exposed occupationsRise up to 27% stronger than in little-exposed occupations among under-50s (Sept. 2024 to Aug. 2025); postings down 38% in relative termsKOF, ETH Zurich, October 2025
Share of the rise in unemployment attributable to AIAt most one fifth of the rise over the past three yearsMichael Siegenthaler (KOF), SRF; Adecco Q3 2026
Highly exposed jobs28% of the 3 million jobs analysed (86 occupations), i.e. over 850,000 jobs; 34% moderately, 38% little exposedeinstAIn, Employés Suisse and Kuble, 16 June 2026
Automatable tasks18% of tasks in Switzerland, 4th of some thirty economies (Luxembourg 21%, United Kingdom nearly 20%)Coface and OEM, 30 June 2026
Entry-level postings-32% in 2025 compared with the 2019-2022 average; 44% of postings are in exposed occupationsjobs.ch and jobup.ch, AI Report, June 2026
Postings requiring AI skills25,000 in 2025, 9,000 more than in 2024, but 1.8% of the marketPwC, AI Jobs Barometer, 15 June 2026
Vacancies in skilled trades and technical occupations+131% between 2015/16 and 2025/26 against +30% for the whole market; 2% of these occupations at high risk of automationAdecco and University of Zurich, 15 October 2026
Worldwide, by 2030170 million jobs created, 92 million displaced, i.e. +78 million net; 39% of skills transformedWEF, Future of Jobs Report, January 2025
Unemployment in Switzerland141,015 unemployed in October 2026, rate of 3.0%, +4.3% year on year; young people 3.2%SECO, 5 November 2026

Three readings. First, exposure is not elimination: the authors at einstAIn, Coface and KOF all repeat it, an automatable task is first of all a redistributed task. Second, the effects are already visible, but concentrated in a few office and IT occupations. Third, Swiss unemployment remains low, and AI explains only a fraction of it: the global economy weighs more.

Why this wave is not like the previous ones

Computers and robots mostly replaced repetitive tasks, often in low- or medium-skilled jobs. Generative AI does the opposite: it writes, codes, classifies and researches. According to the exposure measure used by KOF, the most affected occupations are coders and proofreaders (index 0.96), application programmers (0.94) and database developers (0.91), while cleaning staff, caretakers and sewers are at zero. In other words, skilled, well-paid and often graduate occupations are on the front line, which the Federal Council also notes in its report of 22 April 2026 on labour shortages.

Speed is the other difference. It took desktop computers decades to leave a trace in unemployment statistics. Language models were adopted within months, without heavy investment, and KOF observes that the unemployment curves of highly and little-exposed occupations, parallel for twelve years, diverge precisely from late 2022. In the last year observed, highly exposed occupations count about 7,400 more unemployed than little-exposed ones.

What the figures do not yet say is the overall effect. KOF measures gaps between groups of occupations, not total employment, which grew by 7.4% between mid-2020 and mid-2025. Productivity gains can create demand elsewhere, and the Federal Council considers it "still too early to assess the long-term effects of AI on the labour market". Its next report on digitalisation is announced for 2027.

The occupations where jobs are already tightening

Four Swiss sources point to the same families of occupations, each with its own method:
01

Software development

This is the clearest case. According to the KOF authors, the number of unemployed developers almost tripled between late 2022 and late 2025, after fifteen years of stability, and postings for highly qualified IT roles fell by 18% in 2025 (Adecco, 29 January 2026). Hiring linked to AI deployment did not offset the decline. Demand picked up slightly in the second quarter of 2026 (+6%), but for experienced profiles.

02

Administration, accounting and sales

Postings for commercial, administrative and sales occupations fell by 20% in 2025, then by a further 13% year on year in the second quarter of 2026 (Adecco). In the einstAIn analysis, almost all the administrative occupations studied are highly exposed, as are 72% of those in finance and law. Coface estimates that more than a quarter of tasks in management, administration, law and finance can be automated.

03

Marketing, communication and HR

KOF notes marked rises in unemployment among marketing specialists, journalists and recruiters. The jobs.ch AI Report describes the same movement with its scarcity index: between 2023 and 2025, AI-exposed roles became 69% easier to fill, and social media managers no longer count among scarce profiles.

04

Tasks, more than occupations

In each family, it is the standardisable tasks that move: data entry, first drafts, reporting, checks, first-level support. Coface describes the first stage as a redistribution of tasks within existing positions, with less routine at junior or back-office level and more weight given to judgement, expertise and the supervision of systems.

05

And wages

The jobs.ch AI Report observes a decline in the mid-to-high salary bands in sales, customer service and administration (-53% share since 2023), while the low-salary segment gains ground. Conversely, medical postings above CHF 135,000 rose from 8% to 32% of very well-paid postings.

Beware of conflation: a job in banking or retail is not a threatened job in itself. Exposure says that part of the tasks can be accelerated. What happens next depends on demand in your sector and on your employer’s choices.

The occupations that are hiring, with or despite AI

The Swiss job market is splitting into two realities, in the words of jobs.ch. Here is the other half:
01

Health and care

Highly qualified health occupations grew by 10% in 2025 and by 17% year on year in the second quarter of 2026 (Adecco), the strongest rise of all groups. According to the jobs.ch AI Report, a nurse is eight times harder to find than in 2023, and Switzerland expects a shortfall of more than 65,000 professionals by 2030. AI helps with diagnosis, but interpretation and decisions remain human.

02

Skilled trades and technical occupations

This is the surprise of the decade. Vacancies in these occupations grew by 131% between 2015/16 and 2025/26, four to five times more than the market: +173% in construction and finishing trades, +134% among electricians and electronics technicians, +127% among mechanics. Only 2% of these occupations are considered at high risk of automation, and 19% of their workers are 55 or older, which heralds replacements (Adecco, 15 October 2026).

03

Field and customer-facing jobs

Hospitality, logistics, production, in-store sales, security: little-exposed roles became 19% harder to fill between 2023 and 2025 (jobs.ch). SECO sums it up: occupations that cannot be automated should gain in importance.

04

Jobs that use AI, not only those that build it

In 2025, 25,000 Swiss postings required AI skills, 9,000 more than in 2024. Jobs that apply AI grew by 61%, those that develop it by 8% (PwC, 15 June 2026). Advertised salaries with AI skills are 35% higher in energy and 30% higher in healthcare. The jobs.ch AI Report confirms that AI is leaving IT: the share of AI-related postings gained 186% in construction, architecture and engineering, and 133% in finance and fiduciary services.

05

AI specialists themselves

About 5,000 AI jobs were posted between July 2025 and June 2026, 32% more than a year earlier, driven by SMEs (76% of the growth) and the public sector (+67%). Zurich accounts for 45% of postings, Geneva around 620 and Bern 590 (HSLU, October 2026). But only 5% are permanent junior positions, and newcomers describe the market as "extremely competitive".

06

Worldwide

The WEF puts big data, fintech, AI and software development specialists at the top of the fastest-growing roles by 2030, along with renewable energy and electric vehicle engineers. Declining: cashiers, administrative assistants, bank tellers and data entry clerks.

For the most sought-after occupations in French-speaking Switzerland, our top 10 most in-demand jobs gives the detail by occupation, and Licorn’s cantonal pages track jobs in Geneva or in your canton month after month.

Starting a career in the age of AI: what the postings show

The first to be affected are not those who have a job, but those looking for one. The jobs.ch AI Report, based on 7.3 million postings published between 2019 and 2025, puts the drop in entry-level postings at 32% in 2025 compared with the 2019 to 2022 average. In AI-exposed roles (administration, HR, banking, finance, marketing, procurement, sales, IT), the share of junior positions fell by 16% and that of senior positions rose by 26%. The tasks once given to newcomers, data entry, first analyses, first drafts, scheduling, are precisely those the tools accelerate.

KOF reaches the same conclusion through unemployment: the rise in exposed occupations is more marked among under-50s, who change jobs more often, and the technology seems to complement experienced people and replace those starting out. The difference by age nevertheless remains less marked than in the United States. In the jobs.ch survey, 41% of under-25s and 37% of 25-34 year olds say they are very worried about one day being partly replaced by AI, French-speaking respondents more so than German-speaking ones.

Two nuances. Youth unemployment in Switzerland stands at 3.2% in October 2026, barely above the general rate, and the economic climate explains a good part of the hiring freeze, as the report’s authors themselves point out. And a US analysis cited by economiesuisse (Ramp and Revelio Labs, over 21,000 companies) shows that companies that integrate AI deeply increase their headcount by about 10% within two years, and their entry-level positions by 12%. It is the companies that use it little that hire little. The guide to jobs after 50 covers the opposite end of the pyramid.

Six moves to position your job search

None of the following guarantees a job. But each move is based on what employers actually ask for in Swiss postings:
01

Look at your tasks, not your job title

List what you do in a week and mark what a tool already does faster. The einstAIn dashboard gives the exposure of 86 Swiss occupational groups. What remains is your value: decisions, clients, the unexpected, responsibilities. That is what belongs in your CV.

02

Put transversal skills forward

In more than a million Swiss postings, the share of transversal skills rose from 60% to over 63% of mentions since 2022, and they appear in 85 to 90% of postings (Adecco, 23 April 2026). The most requested: initiative, teamwork, autonomy, willingness to learn. Employers ask for fewer diplomas and more transferable abilities.

03

Use AI in your work, and say so

31% of the population already uses it for professional purposes (FSO). PwC shows that companies look first for people who apply AI to their job, not developers, and that these profiles command a wage premium. An application that shows concrete use, a measured time saving, an improved process, speaks to a recruiter. Our guide to AI in recruitment explains how to do it without falling into the generic application.

04

Train continuously, not once

The WEF estimates that 39% of current skills will be transformed by 2030. 57% of Swiss workers surveyed by jobs.ch already train on their own, often informally. The Swiss system is permeable: higher vocational education, CAS, federal diplomas, and a free career check-up from age 40 in most cantons via viamia. The Federal Council explicitly relies on continuing education as the first response.

05

Do not rule out hands-on occupations

If you are hesitating over a change of direction, the figures favour health, technical and construction occupations, where demand is strong and the risk of automation low. SERI notes that AI, by taking over routine cognitive tasks, relatively revalues manual skills. Our guide to succeeding in your career change details the steps.

06

Do not compete with AI on its tasks

This is the common advice from Coface and KOF: value is shifting to judgement, responsibility, relationships and the supervision of systems. The tasks that tools master better and better are not ground to fight on, they are ground to delegate.

The Licorn method: proposals based on your declared skills

When tasks change faster than job titles, keyword sorting misses the point. Licorn works differently:
01

You declare your skills and your requirements

The recruiter declares the position and its conditions. The matching compares criteria set on both sides, salary, workload, location, remote work, skills, not a guessed PDF. You only receive proposals that respect what you have set.

02

A CV readable by humans and machines

Licorn generates your CV in a standard format that any recruiter and any software can read effortlessly. Your real tasks and transversal skills have their place in it.

03

You stay visible without searching

One candidate in five on Licorn is on watch, employed, and sends no applications (Licorn barometer, 19 October 2026). The profile remains invisible to the current employer. When your occupation is shifting, it is the least risky way to see what the market offers you.

FAQ

Frequently asked questions about AI and jobs in Switzerland

In short

In Switzerland, AI has not yet eliminated jobs on a large scale, but it has already shifted demand: fewer postings and more unemployed in software development, administration, marketing and sales, fewer entry doors for newcomers, and a shortage moving towards health, construction, technical and field occupations. Employers ask for fewer diplomas and more transferable abilities, and reward those who use AI in their occupation. Describe your real tasks, show what you can do with and without the tools, and stay visible even while employed: that is what Licorn makes possible with a profile based on your skills and your requirements.

Sources