Swiss made: The Alpine nation’s affair with AI

Examining the state of artificial intelligence and machine learning in one of the world's most advanced countries

Switzerland is renowned worldwide for its delectable chocolate, mouth-watering cheese fondue, and state-of-the-art Swiss army knives. However, relatively few realize that Switzerland is also emerging as an AI powerhouse in the 21st century.

According to the Tortoise Global AI Index 2023, Switzerland ranks ninth overall and third in Intensity, following Singapore and Israel, indicating the nation’s AI power relative to its size. In particular, Switzerland excels when it comes to three constituent submetrics: Research and Development, where it ranks fourth globally, and Commercial Implementation as well as Talent, where it’s ninth.

However, the country ranks relatively low in terms of Scale — at number 16 — which measures its absolute AI capacity. The submetrics of Operating Environment and Government Strategies are particularly unimpressive, placing the Alpine nation 30th and 56th in the world, respectively.

There seems to be some sort of disconnect... But what is it? Let’s find out together. 

Promising figures

First, let’s look at what makes Switzerland stand out. According to this year’s market insights by Statista, things are really looking up for the Western European country of 9 million people. 

In terms of GenAI, Switzerland’s market size is projected to reach $650 million by the end of this year. Furthermore, it’s expected to grow at an annual rate of over 46%, with the market volume reaching $6.40 billion by 2030.

Image by Statista

In terms of AI more broadly, the market size is projected to reach $1.74 billion by the end of this year. And with an estimated annual growth rate of over 28%, the market’s volume is expected to reach $7.71 billion by 2030. 

Notice how in the picture below, the Swiss AI market today is made up almost equally of computer vision, AI robotics, and autonomous sensory technology. This is quite telling, and it’ll become obvious why a little later.

Image by Statista

Proficiency = growth

According to last year’s figures from OECD, almost 56% of all AI professionals in Switzerland are under the age of 35 and with postgraduate degrees.

Image by OECD

These figures go hand in hand with last year’s survey by Deloitte, as they do with this year’s Stanford AI Index Report. After interviewing over 1,000 Swiss individuals working with computers in a professional setting, Deloitte reports that 73% use AI tools for two hours a day on average, while 27% use them for longer periods. In line with the global trend, 61% of all those surveyed utilize GenAI on a regular basis.

Image by Deloitte 

At the same time, Stanford’s recent data indicates that Switzerland is not only highly AI-proficient on the end-user’s end, but on the developer’s end as well. In fact, Switzerland is second only to Luxembourg globally in attracting AI talent through migration.

Significantly, Stanford reports a 117% increase in IT and Computer Science bachelor’s degree graduates in Switzerland over the past decade and a 148% increase in master’s degree graduates over the same period, placing Switzerland among the top five nations in AI education.

“All your base are belong to us” 

Stanford’s research also offers us a glimpse into the state of Swiss AI within the commercial sector. Over 120 newly funded AI companies were established in Switzerland between 2013 and 2023. Even more strikingly, as of 2022, there were at least 35 professional robot manufacturers in Switzerland.

According to some sources, Switzerland is now a global leader in AI drones. The so-called “Drone Valley,” located between Lausanne and Zurich, boasts over 80 drone manufacturers employing over 2,500 staff. Meanwhile, the Greater Zurich area is known as “the Silicon Valley of robotics,” where companies routinely produce AI robots for a wide range of applications. 

Many of these firms have offices in the US and serve both domestic and international clients. Among them are Fotokite (first responder drones), Verity (warehouse and inventory drones), Wingtra (survey data drones), ANYbotics (inspection drones), Voliro (drones for several key industries), aiEndoscopic (medical robots for endoscopy), and Sevensense (computer vision and sensory solutions for AI robots). 

Notice how these facts align neatly with the figures from Statista, indicating that Switzerland excels precisely in the AI industry subcategories crucial for developing and testing autonomous AI robots, namely computer vision and sensory technology.

A finger in every pie 

Notably, it’s not just the private sector in Switzerland that’s invested in AI robotics — it’s also the government, and a foreign one at that. DARPA (the Defense Advanced Research Projects Agency), part of the US Department of Defense, funds a regular challenge aimed at developing innovative robotic technologies, where Swiss companies play a vital role. 

Upon closer examination, it becomes clear that these AI robotic companies are acting as direct facilitators in Swiss-American relations. The whole challenge effectively serves as a testing ground for the American government, enabling Washington to familiarize itself with and potentially acquire the latest and most advanced AI technologies for use in rescue or even combat operations.

Scientia potentia est

One standout team from ETH Zurich frequently outperforms most others in DARPA’s challenge, which comes as no surprise given the institution’s status as a highly prestigious global university, ranked 7th overall and also in AI, with a competitive Master’s degree in Data Science.

AI research in Switzerland extends to the EPFL AI Center, with over 40 labs and 400 researchers studying intelligent systems. CAIM, established by the Bern University Hospital, advances AI in healthcare. CSEM collaborates with the University of Geneva and University of Neuchâtel to create AI solutions for the industrial sector. iCoSys at the School of Engineering and Architecture in Fribourg focuses on distributed computing and smart living.

Impressively, CSCS houses ALPS — one of the world’s “most AI-capable” supercomputers with over 10,000 GPUs of the new NVIDIA Grace Hopper superchip. Idiap, a semi-private institute with national importance, develops tools for detecting deepfakes together with PXL Vision. Meanwhile, the Swiss AI Initiative unites these centers under a single umbrella, overseeing joint, often multidisciplinary, AI projects across the nation.

Muscles out of sync

Regarding concrete AI policies by the Swiss government, the situation is less rousing — there aren’t many of them. This reflects Switzerland’s low Government Strategies metric, which negatively affects its overall Scale score.

The Digital Switzerland initiative, introduced in 2018, established a working group within the federal administration led by the State Secretariat for Innovation (SERI). This includes creating a network of AI-enabled smart cities, but despite noticeable progress, the plan is still far away from being fully realized.

The more recent AI Action Plan, announced by the government in 2020, consists of seven guidelines — from “putting people first” to “involving all relevant stakeholders.” While it emphasizes AI accountability and safety, it lacks detailed implementation strategies, serving more as an ideological framework.

Not everything is ideal within the commercial sector either, contributing to the country’s low Operating Environment score, which is another submetric of Scale.

According to a new Amazon Web Services study, Swiss businesses experienced a 36% increase in AI adoption last year. In line with Deloitte’s findings, nearly half of Swiss companies today use AI to optimize business processes. Nonetheless, over 60% of Swiss firms also note a limited choice of AI providers and models, hindering full AI integration. Quite revealingly, this voiced concern is nearly 40% higher than the European average.

Conclusion

Although Switzerland is a prominent AI force, its AI implementation is not yet well balanced. Many top-class Swiss AI companies exist, but most of them focus on a select few sub-industries, primarily robotics. The public sector shows even less variability, with most government initiatives still in their nascent stages.

Despite numerous high-level AI facilities and substantial market growth, Switzerland ultimately falls short of providing an all-encompassing environment for AI development. While there’s significant activity (hence the high Intensity score), it’s all rather one-sided (hence the low Scale score), reflecting the nature of the Swiss economy in general. Swiss AI companies tend to specialize in a limited number of niche areas rather than being evenly distributed across various sectors.

With that said, the situation may change as incoming talent brings new ideas and opportunities, leading to a broader range of AI offerings. In parallel, as research centers demonstrate flexibility and move beyond their “comfort zones,” the smart city initiative could be realized sooner, improving the nation’s overall AI standing. CERN, renowned for particle physics and now beginning to lean towards AI, is a prime example of this propitious pattern.

With more variability introduced into Switzerland’s pro-innovation environment, reinforced by new faces and businesses of all sizes gaining access to newer home-grown solutions, the nation’s absolute AI capacity is likely to rise.