How AI could reshape Switzerland’s labour market

AI as a productivity lever

Demographic change is placing increasing pressure on the Swiss labour market. Over the next ten years, more than 900,000 employees will reach retirement age and leave the workforce. This represents around 20 per cent of today’s working population. As a result, Switzerland faces substantial demographic replacement demand, which could further intensify existing labour shortages across a wide range of occupations and industries.

This replacement demand will not be distributed evenly across the labour market. Some occupational fields already have low unemployment combined with strong demand for skilled workers. If many employees in these fields retire over the coming decade, labour shortages could become particularly acute. Other occupations face higher unemployment, weaker demand for labour or comparatively moderate replacement needs. The key issue is therefore not simply how many people will leave the workforce, but also which occupations and industries will be most strongly affected.

Productivity growth via use of technology has long been one of the most important responses to labour shortages. Computers, automation, mobile technologies and digital platforms have enabled businesses to produce more with less labour, while also creating new business models and new types of work. The question now is: could artificial intelligence become the next major productivity lever, and if so, could it help Switzerland relieve some of the demographic pressure it faces?

The answer to that question is that although AI cannot eliminate the problem entirely, it will certainly be able to help alleviate future labour market pressures. However, its impact will not be spread evenly across the economy.

Where AI meets demographic change

To assess the potential of AI in an ageing society, we combined Swiss occupational and demographic data with international evidence on how generative AI is being used across occupational tasks. We first estimated demographic replacement demand by analysing how many people in each occupation are likely to retire within the next ten years. We then compared this with our newly developed AI Hotness Index, which captures both current AI use and the potential for AI to support or automate occupational tasks. Together, these two dimensions show where AI could help alleviate workforce pressures, where its immediate impact is likely to remain limited and where it may fundamentally reshape existing roles.

Our analysis reveals that demographic replacement demand cuts across occupations with very different skill profiles and areas of work. The highest replacement rates are found across professional and public-sector occupations, transport and skilled trades, as well as care and service roles. Judges, for example, face the greatest replacement need, while bus and tram drivers, toolmakers and domestic cleaners also rank highly. Significant replacement demand is also evident in education, health care, management and administrative occupations. In total, over 900,000 workers are set to retire within the next ten years, which represents around 20% of today’s workforce. Among the 279 occupations analysed, 109 have replacement demand exceeding this average over the coming decade.

AI usage, by contrast, is much more concentrated in occupations involving digital, information processing and communication-intensive tasks. As a result, ICT and software-related occupations dominate the top of the AI Hotness ranking followed by a range of language and content-related occupations, such as translators, authors and journalists. Administrative and communication roles including secretaries, office clerks and public relations professionals are also seeing high levels of AI exposure. These occupations share a high proportion of cognitive and digital tasks that can already be supported by today’s AI systems.

Top and bottom 5 occupations by AI Hotness and replacement demand

AI Hotness

Analysing current patterns of AI usage provides valuable insight into where AI is already influencing work today. To do so, we created an AI Hotness Index that combines two dimensions:

• Actual AI usage data based on Claude and Copilot conversations from Anthropic and Microsoft, respectively, mapped to occupational tasks.
• Exposure estimates developed by OpenAI researchers and collaborators that assess the potential for LLMs to reduce the time required to perform.

Replacement demand

The replacement demand is calculated by taking the share of people aged 55 and over per occupation. It highlights where retirements are likely to create the greatest hiring needs over the coming decade. However, it reflects the economy as it exists today and is therefore a static indicator that does not account for future changes.

Comparing the demographic replacement demand indicator with the AI Hotness Index shows that AI will not fundamentally solve the challenges posed by demographic change. However, it will still have an impact which will vary considerably across tasks and occupations.

Our analysis reveals that AI will not uniformly solve Switzerland's demographic labour challenge. In knowledge-intensive occupations, AI can meaningfully offset replacement demand. In care, trade and people-centred occupations, labour shortages will persist. And in occupations already facing weaker demand, AI may accelerate the need for workforce transitions. 

This differentiated picture is essential, not only for policymakers and business leaders, but equally for individuals. A one-size-fits-all approach will not work. Instead, companies and government must pursue targeted strategies: immigration for some sectors, training and upskilling for others, productivity investment for still others. But success ultimately depends on individuals taking responsibility for their own careers. 

Workers must embrace lifelong learning, remain open to technological change and be willing to adapt their skills as occupations evolve. The responsibility for managing these transitions is shared across three levels: policymakers must create the framework and invest in education; companies must invest in their workforce and adapt their business models; and individuals must take ownership of their own development and remain agile in the face of change.

Sample job profiles

What corporates need to do

AI and demographic change must be managed together. AI cannot reverse workforce ageing, but it can help scarce workers handle larger workloads, automate some tasks and reshape roles, while having limited impact in others. Companies should combine the quadrant logic with their own evidence on recruitment conditions, expected demand, workforce structure and processes.

We propose a response in three layers: