Swiss SMEs embrace AI while expecting limited job losses

Seventy-four per cent of Swiss SMEs now use artificial intelligence every day, according to an Axa study published on September 15, 2026. Adoption accelerated sharply between 2024 and 2025, rising from 55% to 74%, and has since stabilised at that level.

The figures place AI inside the ordinary routines of Switzerland’s small and medium-sized businesses, rather than at the margins of corporate experimentation. The technology is being used across daily operations, with companies extending its role beyond basic text-based tasks.

That spread matters in an economy where SMEs account for much of the country’s business activity and employment. Yet the study does not describe a labour market shock. Most firms expect AI to make existing work more efficient, while only a small minority foresee direct job losses.

The findings offer a measured picture of adoption. Swiss businesses are incorporating AI while retaining familiar organisational structures and cautious expectations about what the technology can deliver. For employees, the immediate story is one of changing tasks and potential time savings, rather than mass displacement.

AI is moving beyond writing and into the machinery of everyday business. The Axa study records a rise in the use of AI for process optimisation, from 23% to 35%. Applications for data analysis also increased, from 22% to 34%.

That shift suggests companies are using the technology to examine information, streamline workflows and support operational decisions. The source does not identify individual firms or sectors, so it offers no basis for ranking which branches are moving fastest. It does show a broader pattern: AI adoption is becoming more embedded in business processes.

For a small company, process optimisation can cover recurring administrative work, while data analysis can help staff sort and interpret large volumes of information. Such uses may reduce time spent on routine tasks without removing the people responsible for checking results, dealing with exceptions or making decisions.

The change also raises practical demands. Employees need to understand how systems produce recommendations, where errors may occur and when human review is necessary. The study’s numbers point to expanding use, although they do not measure productivity gains or the quality of individual AI systems.

Customer-facing AI remains the cautious frontier for Swiss SMEs. The study says applications such as personalised advertising and customer service are stagnating, with trust in AI during direct customer contact still limited.

That hesitation reflects the risks of placing an automated system between a company and its customers. A flawed answer can damage confidence quickly, particularly for businesses whose reputation depends on personal service, specialist advice or local relationships. The source does not give a breakdown by canton, industry or company size, so it cannot show whether this pattern is stronger in one part of Switzerland than another.

Internal applications carry a different risk profile. Staff can review an analysis, correct a draft or stop a process before it reaches a customer. Direct interaction leaves less room for recovery when an answer is inaccurate, impersonal or poorly timed.

For now, Swiss SMEs appear more comfortable using AI behind the scenes than presenting it as the voice of the business. That divide helps explain why overall adoption can be high while customer-facing deployment remains restrained.