Whoever imports AI exports value creation. Whoever exports AI imports prosperity.
Within five years, build a national AI competence centre that links ETH, EPFL, IDSIA and the universities of applied sciences with industry. At least 500 million francs in public funding, co-financed by the private sector.
Switzerland has an AI landscape that far exceeds its size:
| Institution | Contribution |
|---|---|
| ETH Zurich | Among the ten best technical universities in the world. ETH AI Center since 2020 in Zurich-Oerlikon [1]. |
| EPFL | Leading in robotics, computational neuroscience and machine learning [2]. |
| IDSIA Lugano | Juergen Schmidhuber and Sepp Hochreiter invented the LSTM network here in 1997 -- a foundational technology of deep learning [3]. |
| Google Zurich | Largest Google site outside the USA, over 5,000 employees [4]. |
| IMF AI Preparedness Index | Switzerland at rank 3 of 186 countries [5]. |
There are many individual initiatives. What is missing is a national system:
A national AI competence centre would bundle the existing strengths:
| Pillar | Participants | Focus |
|---|---|---|
| Fundamental research | ETH, EPFL, IDSIA | New algorithms, safety research |
| Applied research | Universities of applied sciences (FHNW, ZHAW, BFH, HES-SO) | Industry solutions, SME transfer |
| Industry partners | Novartis, Roche, ABB, UBS, Swisscom, start-ups | Funding, use cases, scaling |
| Education | All universities + vocational colleges | AI courses, continuing education, retraining |
| Item | Amount | Context |
|---|---|---|
| Public funding (Confederation) | CHF 500 million over 5 years | CHF 100 million/year = 0.013% of GDP |
| Private sector co-financing | at least CHF 500 million | |
| Total | at least CHF 1 billion | 0.025% of GDP over 5 years |
Switzerland's gross domestic product (GDP) stood at around 813 billion francs in 2024 [6]. The requested 500 million over five years amounts to 100 million per year -- that is 0.013 per cent of GDP. For comparison:
| Item | Amount/Year | Share of GDP |
|---|---|---|
| Requested AI investment | CHF 100 million | 0.013% |
| ETH domain (total budget) | CHF 3,500 million | 0.43% |
| National defence | CHF 6,000 million | 0.74% |
| Road infrastructure (FEDRO) | CHF 4,200 million | 0.52% |
| Looming fiscal gap from AI | CHF 75,000--116,000 million | 9--14% |
The calculation is sobering: Switzerland risks an annual fiscal gap of 75 to 116 billion francs -- and is supposed to invest 100 million per year to counter it. The ratio is 1:750 to 1:1,160. Put differently: for every franc that the AI transformation could cost the state, 0.1 centimes is invested in preparation.
The 500 million is not an ambitious programme -- it is the absolute minimum. Measured against what is at stake, the demand is not too high but possibly too low [7].
| Country | Investment | Result |
|---|---|---|
| Singapore | USD 750 million for national AI strategy | AI hub for South-East Asia [7] |
| Israel | USD 300+ million annually, plus military spin-offs | "Start-up nation", leading in AI security [8] |
| South Korea | USD 2 billion for AI by 2029 | Samsung, LG, Hyundai as AI drivers [9] |
The watch industry lost over 60 per cent of its jobs in the quartz crisis of the 1970s and 80s. Watchmakers in the Jura who had assembled mechanical watches found themselves competing against Casio quartz watches for ten dollars. Written off. Then came Swatch, then the renaissance of luxury watches. Today the Swiss watch industry is stronger than ever -- because it bet on innovation and quality, not on protectionism [10].
The pharmaceutical industry has proved that fundamental research pays off. Basel is the world capital of pharmaceuticals because Roche and Novartis invested in research for decades.
The same logic applies to AI -- only the timescale is shorter and the stakes are higher. Every franc invested here is not a cost item -- it is an insurance premium against economic decline.
[1] ETH AI Center, Zurich-Oerlikon.
[2] EPFL, AI Center and School of Computer and Communication Sciences.
[3] Hochreiter, S. / Schmidhuber, J.: Long Short-Term Memory. In: Neural Computation 9(8), 1997.
[4] Google Switzerland: Zurich location, over 5,000 employees.
[5] IMF, AI Preparedness Index, 2024.
[6] ETH Board: Strategic planning and budget for the ETH domain.
[7] Singapore National AI Strategy 2.0, 2023.
[8] Israel Innovation Authority: National AI Program.
[9] Republic of Korea, Ministry of Science and ICT: AI Strategy, 2024.
[10] Federation of the Swiss Watch Industry: Annual reports.