TLDR: Swiss employers now hire far more people who use artificial intelligence than people who build it, so the winning profile pairs a core profession with documented, everyday fluency.
PwC counted about 25,000 Swiss job ads requiring artificial intelligence skills in 2025, most of them for users
PwC Switzerland’s 2026 AI Jobs Barometer, published on 15 June 2026, counts about 25,000 Swiss job advertisements that required artificial intelligence (AI) skills in 2025, a record and roughly 9,000 more than a year earlier. Inside that total, advertisements for AI user profiles, people who apply AI tools inside an existing job, rose by about 8,400 to a new high. Advertisements for AI developer profiles, people who build models and systems, rose by about 220. On those two figures, Swiss employers created roughly 38 new user roles for every new builder role.
PwC’s Switzerland country analysis, built on Lightcast job-posting data for 2018 to 2025, adds a second layer: AI user roles are rising above their 2022 peak, while developer roles remain below theirs. Edward Galle’s earlier analysis Data Scientists, Engineers and Analysts Fill 43% of Swiss AI Job Ads covers the builder side of the market, where data engineering carries most of the hiring. This article covers the far larger user side, and the practical question it raises for candidates and employers in regulated functions: how to document everyday AI use so that a hiring manager, a reviewer or a supervisor can rely on it.
PwC also keeps the trend in proportion: “Nur 1,8 % aller Stelleninserate in der Schweiz haben KI-Bezug”, meaning only 1.8% of all Swiss job advertisements refer to AI. Read from the hiring side, that small share is an opportunity. The profiles employers explicitly request remain scarce, and professionals who can document their AI use stand out in a market where most curricula vitae (CVs) still describe it in a single vague line.
| Swiss artificial intelligence (AI) hiring indicator | 2025 value |
|---|---|
| Job ads requiring AI skills | about 25,000 (about +9,000) |
| Share of all Swiss job ads | 1.8% |
| Change in AI user profiles | about +8,400 |
| Change in AI developer profiles | about +220 |
| User share of AI roles, financial services | 95.9% |
| Developer share of AI roles, technology, media and telecoms | 16.2% (highest) |
| Average new skills per occupation, most AI-exposed quartile | 248 |
Financial services and health show where documented artificial intelligence use matters most
In financial services, PwC reports that 95.9% of AI-related roles are user profiles, the highest user share of any Swiss sector. Banks, insurers and asset managers largely buy, configure and supervise AI tools, and the jobs they advertise are credit analysts, compliance officers and client advisers who use those tools inside a supervised process. The technology, media and telecoms sector sits at the other end, with the highest developer share at 16.2% of its AI advertisements, and even there users dominate.
The Swiss Financial Market Supervisory Authority (FINMA) surveyed about 400 licensed institutions and reported in April 2025 that around 50% of the institutions surveyed use AI or have initial applications in development, and that 91% of those using AI also use generative AI. A single Swiss bank job advertisement shows how that reaches hiring. Swissinfo reported on 22 July 2026 that Sygnum advertised a communications manager role under the headline “AI-native candidates preferred”, stating: “At a minimum, you must have hands-on experience using a range of AI tools in your daily work.” The same report cites the Swiss Employees’ Association estimate that more than CHF 80 billion of Swiss wages are now directly linked to workers who can use AI for their employers’ benefit.
PwC also reports that AI skills command above-average salaries in Switzerland, with the largest premiums in health and energy. Health matters for the life-sciences market Edward Galle serves: the premium appears precisely where regulation is heaviest and where an employer pays for the rare combination of domain judgement and tool fluency. Edward Galle’s analysis Digital Hires in Swiss Pharma Are Priced Against the Technology Market, Above the Pharma Median gives the salary baseline against which that premium applies.
AI fluency grows inside existing jobs as a skill layer on each profession
The PwC country analysis also tempers the popular story of an AI jobs boom: since 2018, job postings in Swiss occupations with the lowest AI exposure grew to around 15.4 postings for every posting in 2018, against 8.7 in the most exposed occupations. Demand for AI capability is rising across every sector, while the occupations growing fastest in volume are those AI touches least. Read together with the user-versus-builder split, the message is consistent: AI in the Swiss labour market behaves as a capability layered onto existing professions, and each profession keeps its own core.
The same analysis shows how fast that layer changes the job underneath. PwC finds the most AI-exposed quartile of occupations averaging 248 new skills per occupation. The job title stays the same while the skill list underneath it is rewritten. A marketing manager, a pharmacovigilance scientist or a financial controller therefore gains more by deepening AI use inside the current discipline than by retraining as a machine-learning engineer. Builder roles remain valuable and well paid, and they remain few. PwC’s own recommendations to Swiss leaders point the same way, stressing human skills such as judgement, empathy, creativity and leadership alongside AI skills.
That reading also explains why documentation decides the value of AI fluency. A skill that sits inside every profession becomes visible only through the work it changed. In regulated functions, the work it changed is the work a reviewer, an auditor or a supervisor will later examine.
Regulated employers already expect artificial intelligence use to be documented
Swiss and European supervisors have set out the same expectation in different words. FINMA’s Guidance 08/2024 of 18 December 2024 describes “a centrally managed inventory with a risk classification and resulting measures”, “the definition of responsibilities and accountabilities for the development, implementation, monitoring and use of AI”, and the need to “assign responsibilities clearly to employees with the appropriate skills and experience”. In life sciences, the European Medicines Agency (EMA) states that its committees adopted a reflection paper on the use of AI in the medicinal product lifecycle in September 2024, setting out how companies should manage AI used anywhere from development to post-authorisation activities.
Hiring follows the same logic: under the European Union (EU) AI Act, as amended by the Digital Omnibus, employers that use AI to screen candidates take on documentation and oversight duties from December 2027. A clear description of a role’s AI tasks feeds directly into that record, so the job advertisement becomes the first page of the governance file.
The Microsoft and LinkedIn Work Trend Index of May 2024 shows training lagging use: in a survey of 31,000 knowledge workers in 31 markets, only 39% of people who use AI at work had received AI training from their company. Most everyday AI use in regulated companies is therefore self-taught, and self-taught use carries weight in a regulated function only when the person can show the task, the tool, the control and the result.
A CV entry for everyday AI use names the task, the tool, the control and the result
Edward Galle’s view, from its search work for Swiss pharmaceutical, medical-technology and financial-services employers, is that a credible CV entry for AI use answers four questions in one or two lines. Which task did the AI support? Which tool or platform category was used, inside which approved environment? Which control kept the output reliable, such as a human review step, a validation check or a sign-off? What measurable result followed? A line that answers all four reads as professional evidence. A line that only lists a product name reads as a hobby.
| Element | What to write | Weak versus strong |
|---|---|---|
| Task | The work step the AI supported | “Uses AI” versus “first-draft case narratives for adverse-event reports” |
| Tool | Platform category and approved environment | “ChatGPT” versus “company-approved generative assistant inside the safety database” |
| Control | The review, validation or sign-off step | “Checked outputs” versus “every draft reviewed and signed by a qualified scientist” |
| Result | A measurable change the candidate can defend | “Saved time” versus “shorter drafting cycle, with the before and after stated” |
In pharmacovigilance, a professional can describe AI-assisted case intake or literature screening and the medical review that follows; Edward Galle’s article A 2025 European Union Rule Puts Signal Detection Inside Drug Companies in Europe describes where that work is growing. A regulatory writer can describe AI-assisted first drafts of summaries and the quality check that releases them. A credit analyst or controller can describe an AI-supported forecast or reconciliation and the four-eyes review that approves it. A marketer in pharma can describe AI-generated content variants and the medical, legal and regulatory review that cleared them, and a human resources partner can describe an AI-assisted workflow in the company’s human resources information system and the safeguard around it.
An honest entry follows three rules: name only tools the employer approved, because a regulated hiring manager will read unapproved use as a compliance risk; describe confidential work at the level of the process, leaving out client, patient or product data; and state results the candidate can explain at interview, with the baseline and the method of measurement ready, since a number stated with its method survives the first hard question.
A job ad for an AI user role in a regulated function names the same four elements
For employers, the PwC data carries a practical lesson about the advertisement itself. When nearly every new AI-related role is a user role, a job advertisement that asks generically for “AI skills” invites the wrong comparison: builders apply for user roles, users undersell themselves, and screening software matches on the phrase. The stronger advertisement in a regulated function uses the same four elements as the strong CV entry. It names the task the AI supports, the tool category or platform in use, the control the person must respect, and the outcome the team expects. “Uses a company-approved generative assistant to draft adverse-event case narratives, each reviewed and signed by a qualified scientist” tells a candidate exactly what the job is and tells the screening tool exactly what to look for.
The same Work Trend Index found that 71% of leaders would rather hire a less experienced candidate with AI skills than a more experienced candidate lacking them. Set against the 39% training figure, that preference points to a choice Swiss employers can make deliberately: hire for the core profession and the habit of learning, document the AI tasks the role will perform, and train for the specific tools. For a role in a regulated function, ten years of pharmacovigilance or treasury experience is harder to buy than six months of tool practice.
The advertisement also opens the employer’s own record: a role description that lists the AI tasks, the approved tools and the review step gives the AI inventory a named owner from the first day, which is exactly the “responsibilities and accountabilities” structure FINMA describes for supervised institutions and a sound pattern for any regulated employer.
Documented AI use becomes evidence at interview and in performance reviews
At interview, a hiring manager in a regulated function can ask the candidate to walk through one documented example built from the same four elements: the task, the tool, the control and the result, followed by one question on what the candidate would do if the AI output were wrong. A candidate who answers with a clear escalation path shows the judgement that PwC’s data says employers pay for.
After hiring, documentation feeds the performance review: Edward Galle has covered how companies that count AI logins in performance reviews measure compliance, while capability shows in documented outcomes. An employee who keeps a short record of AI-supported tasks, controls and results arrives at each review with evidence, and an employer who asks for that record builds its AI inventory from the bottom up.
Candidates and employers who document everyday AI use turn a common skill into a hiring advantage
Professionals in pharma, medical technology and financial services should rewrite one section of their CV around two or three documented AI entries that state the task, the tool, the control and the result. That evidence shows exactly what PwC’s data says employers now request, and it highlights the review step where generalists fall short. Edward Galle’s overview of the digital skills commercial pharma teams need lists adjacent capabilities that strengthen such a profile.
Employers writing job advertisements should decide first whether the role uses AI or builds it, then describe the AI tasks, approved tools and controls in plain operational language, and budget for tool training. The narrower and more precise the description, the smaller and better the shortlist, and the stronger the first entry in the employer’s AI governance file.
Professionals ready to position that combination of domain depth and documented AI fluency can submit a CV to Edward Galle for confidential matching, or browse current openings in digital, brand-tech and commercial roles across Switzerland.
This article has been reviewed by Haider Alleg, Managing Partner at Kainjoo SA.
References
- PwC Switzerland. AI Jobs Barometer 2026 (press release). 15 June 2026. https://www.pwc.ch/de/presse/ai-jobs-barometer-2026.html
- PwC. 2026 Global AI Jobs Barometer: Switzerland analysis. June 2026. https://www.pwc.com/gx/en/issues/artificial-intelligence/job-barometer/aijb-2026-switzerland.pdf
- PwC Switzerland. 2026 Global AI Jobs Barometer: Swiss cut. 17 June 2026. https://www.pwc.ch/en/insights/transformation/2026-ai-global-jobs-barometer-swiss-cut.html
- Swiss Financial Market Supervisory Authority (FINMA). FINMA survey: artificial intelligence gaining traction at Swiss financial institutions. 24 April 2025. https://www.finma.ch/en/news/2025/04/20250424-mm-umfrage-ki/
- Swiss Financial Market Supervisory Authority (FINMA). FINMA Guidance 08/2024: Governance and risk management when using artificial intelligence. 18 December 2024. https://www.finma.ch/en/~/media/finma/dokumente/dokumentencenter/myfinma/4dokumentation/finma-aufsichtsmitteilungen/20241218-finma-aufsichtsmitteilung-08-2024.pdf
- European Medicines Agency (EMA). Artificial intelligence (reflection paper on the use of AI in the medicinal product lifecycle, adopted September 2024). https://www.ema.europa.eu/ga/about-us/how-we-work/data-regulation-big-data-other-sources/artificial-intelligence
- European Parliament and Council. Regulation (EU) 2026/1744 of 8 July 2026 (Digital Omnibus on AI). In force 27 July 2026. https://eur-lex.europa.eu/eli/reg/2026/1744/oj
- Allen, M. How artificial intelligence is shaping the Swiss job market. SWI swissinfo.ch, 22 July 2026. https://www.swissinfo.ch/eng/swiss-ai/how-artificial-intelligence-is-shaping-the-swiss-job-market/91750448
- Microsoft and LinkedIn. 2024 Work Trend Index Annual Report: AI at Work Is Here. Now Comes the Hard Part. 8 May 2024. https://www.microsoft.com/en-us/worklab/work-trend-index/ai-at-work-is-here-now-comes-the-hard-part
