The academic foundation: ETH Zurich and the University of Zurich

The ETH AI Center

Any account of artificial intelligence in Zurich begins at ETH Zurich, the federal institute of technology founded in 1855, whose computer science department has supplied the city’s technology employers with graduates and researchers for decades. For a long time the university’s AI research was spread across many chairs in computer science, electrical engineering, mechanical engineering and mathematics. In October 2020 ETH brought these activities together in the ETH AI Center, a cross-departmental research centre whose core team consisted of 29 professorships from seven departments. The centre was chaired from the outset by Andreas Krause, professor at the Institute for Machine Learning, while its executive director, Alexander Ilic, came from an entrepreneurial background that included Magic Leap Switzerland. Its programme named machine learning, perception and natural language understanding as research priorities and emphasised reliable, explainable and fair AI, with applications in health, mobility, architecture, energy and climate.¹

The design of the centre reflected a lesson that ETH had learned in its earlier partnerships with industry. Research groups that remain in their home departments are more likely to keep their academic independence, whereas a common centre can offer shared doctoral fellowships, an incubator and a single point of contact for companies. In practice the ETH AI Center became the address through which banks, insurers and technology firms approach the university’s AI researchers. The centre’s current leadership consists of Alexander Ilic as co-founder and executive director, Melanie Gabriel as co-director and chief operating officer and Andreas Krause as chair of the steering committee.² Its fellows, many of whom come from abroad, form one of the most visible talent pipelines between the university and the city’s employers.

The Swiss National AI Institute, the Swiss AI Initiative and Apertus

The second stage in the institutional development was national. In December 2023 ETH Zurich and EPFL in Lausanne launched the Swiss AI Initiative, a collaborative programme with researchers from more than ten Swiss academic institutions. It was seeded with ten million GPU hours on the new supercomputer of the Swiss National Supercomputing Centre so that Swiss researchers could train large foundation models of their own. In October 2024 the two federal institutes announced the Swiss National AI Institute (SNAI), co-directed by the executive directors of the ETH and EPFL AI centres. The ETH Board allocated CHF 20 million for the years 2025 to 2028, to which the two institutions and third parties added further funds. More than seventy AI-focused professors across Switzerland were involved from the outset.³

The Swiss AI Initiative describes itself as the largest open science and open source effort for AI foundation models worldwide, with more than 800 researchers. It operates as the first initiative of SNAI, awards compute grants in periodic calls and counts Swisscom among its strategic partners.⁴ The claim to be the largest effort of its kind is the initiative’s own, yet the numbers of researchers and GPU hours show that Switzerland has chosen to build shared public infrastructure in a field otherwise dominated by a handful of private laboratories in the United States and China.

The most visible result is Apertus, a large language model released on 2 September 2025 in two sizes with 8 billion and 70 billion parameters. It was trained on about 15 trillion tokens of text in more than a thousand languages, with around forty per cent of the training data in languages other than English. The model, its weights, its training data and its training recipes were published under a permissive open-source licence that allows research, education and commercial use. According to the developers, Apertus was built with due consideration to Swiss data protection law, Swiss copyright law and the transparency obligations of the EU AI Act. Swisscom offers access through a sovereign Swiss AI platform, while Hugging Face and the Public AI network distribute the model internationally.⁵

For the financial centre, the main significance of Apertus lies in its function as a proof of concept. A bank that wants to run a language model entirely within its own infrastructure, without sending client data to a foreign provider, can use an open model whose training data and licence terms are documented. The question of whether an open model of this size meets the performance requirements of a particular use case must be answered case by case. What the release establishes is that a sovereign alternative exists and that its provenance can be audited, which is a point of real value for institutions bound by banking secrecy and by FINMA’s expectations on outsourcing.

The Alps supercomputer at CSCS in Lugano

Training models of this kind requires computing power on a scale that universities rarely command. Switzerland’s answer is Alps, the supercomputer of the Swiss National Supercomputing Centre (CSCS), which is part of ETH Zurich but located in Lugano in the canton of Ticino. The system was inaugurated on 14 September 2024 in the presence of Federal Councillor Guy Parmelin. Built by Hewlett Packard Enterprise on the Cray EX platform, it contains 10,752 NVIDIA Grace Hopper superchips and ranked sixth on the TOP500 list of June 2024. Its fully expanded performance was stated at about half an exaflop. Besides AI research, Alps supports MeteoSwiss weather forecasting and scientific computing in many disciplines.⁶

The location of the machine in Lugano illustrates a feature of Swiss AI policy that visitors from centralised states often find surprising. Research is concentrated in Zurich and Lausanne, while the computing centre sits in the Italian-speaking south, and the governance runs through the ETH Domain with its board in Bern and Zurich. For the financial centre, the relevant point is practical: Swiss researchers and their industry partners have access to domestic computing capacity, so that sensitive research can be conducted within Swiss jurisdiction.

The University of Zurich and the rankings

The University of Zurich (UZH), the largest university in Switzerland, complements ETH with strengths in medicine, law, economics and the humanities. Its Department of Informatics hosts groups with international standing, among them the Robotics and Perception Group led by Davide Scaramuzza, which produced the first AI-driven drone to beat a human world champion in drone racing, a result published in Nature. The group belongs to the Department of Informatics and to the Institute of Neuroinformatics, which UZH runs jointly with ETH.⁷ UZH has also institutionalised its reflection on the social effects of technology in the Digital Society Initiative, which describes itself as the university’s competence centre for shaping the digital transformation of society and science.⁸

In 2025 the university’s Digital Strategy Board launched UZH.ai, a hub that coordinates AI activities across more than seven faculties, including law, medicine, economics and linguistics. Its guiding statement, “AI is not the focus — people are”, signals a humanistic orientation that differs from the engineering emphasis at ETH.⁹ For the legal profession and for compliance functions in banks, the combination is useful because questions of liability, explainability and data protection are researched at UZH while the underlying models are developed a few hundred metres away at ETH.

International rankings confirm the standing of the Zurich institutions, although they measure overall reputation and citations and say little about the specific quality of AI research. In the QS World University Rankings 2026, published in June 2025, ETH Zurich held seventh place for the third consecutive year and was the only continental European university in the global top ten, while UZH climbed to 100th place.¹⁰ In the 2027 edition, published in June 2026, ETH shared eighth place with University College London and remained the highest-ranked university in continental Europe. EPFL was ranked 22nd and UZH 98th.¹¹ At the level of the country, the World Intellectual Property Organization’s Global Innovation Index placed Switzerland first in 2025 for the fifteenth consecutive year, with particular strength in knowledge and technology outputs.¹²

Notes

ETH Zurich, “New centre for AI research”, press release of 20 October 2020, https://ethz.ch/en/news-and-events/eth-news/news/2020/10/pr-new-centre-for-ai-research.html. The release also mentions an incubator for AI start-ups, an Open Lab for international collaboration and membership of the European research network ELLIS. ↩
ETH AI Center, “About us”, https://ai.ethz.ch/about-us.html (accessed September 2026). ↩
ETH Zurich, “ETH Zurich and EPFL enhance collaboration to boost AI in Switzerland”, 3 October 2024, https://ethz.ch/en/news-and-events/eth-news/news/2024/10/eth-zurich-and-epfl-enhance-collaboration-to-boost-ai-in-switzerland.html. The co-directors named are Alexander Ilic (ETH AI Center) and Scarlet Schwiderski-Grosche (EPFL AI Center). ↩
Swiss AI Initiative, home page, https://www.swiss-ai.org/ (accessed September 2026). ↩
ETH Zurich, “Apertus: a fully open, transparent, multilingual language model”, press release of 2 September 2025, https://ethz.ch/en/news-and-events/eth-news/news/2025/09/press-release-apertus-a-fully-open-transparent-multilingual-language-model.html. Imanol Schlag (ETH Zurich) is quoted as saying that the model “is built for the public good”. ↩
Swiss National Supercomputing Centre (CSCS), “New research infrastructure: ‘Alps’ supercomputer inaugurated”, 14 September 2024, https://www.cscs.ch/publications/news/2024/new-research-infrastructure-alps-supercomputer-inaugurated. Parmelin is quoted as calling Alps “an expression of our vision of a future characterised by knowledge and progress”. ↩
University of Zurich, Robotics and Perception Group, home page, https://rpg.ifi.uzh.ch/ (accessed September 2026). ↩
University of Zurich, Digital Society Initiative, home page, https://www.dsi.uzh.ch/en.html (accessed September 2026). ↩
University of Zurich, UZH.ai, “About”, https://www.ai.uzh.ch/en/about.html (accessed September 2026). ↩
ETH Zurich, “QS rankings: ETH Zurich secures 7th place once again”, 19 June 2025, https://ethz.ch/en/news-and-events/eth-news/news/2025/06/qs-rankings-eth-zurich-secures-7th-place-once-again.html. ↩
ETH Zurich, “Global QS ranking: ETH Zurich excels in employability, research, and reputation”, 18 June 2026, https://ethz.ch/en/news-and-events/eth-news/news/2026/06/global-qs-ranking.html. ↩
KMU-Portal of the Swiss Confederation, “Switzerland remains the world’s most innovative country”, 8 October 2025, https://www.kmu.admin.ch/kmu/en/home/new/news/2025/switzerland-world-most-innovative-country.html. ↩

For general information only. This article does not constitute legal or tax advice, does not replace advice on an individual case and creates no lawyer-client relationship. It reflects the legal position as of September 2025.

English

Contact

Dr Alexander Ressos, TEP

Rechtsanwalt/Attorney-at-Law

I am a Swiss attorney-at-law and also admitted as a solicitor in Hong Kong and in England & Wales, with almost 25 years of experience.

Together with my multilingual team, including Chinese, Japanese, Korean, Russian, Ukrainian and French native speakers, I advise companies and their owners from around the world on the full range of banking, corporate, commercial and tax law matters in Switzerland, including international tax planning, fiduciary services, and legal family office services including immigration & relocation or real estate investments. My practice has a particular focus on clients from growth and frontier markets in Greater China, Central Asia and the Caucasus, the rest of Asia, the Middle East, Africa and Latin America, and I equally advise clients from Europe and North America. Through an affiliated Swiss trust company, we also take care of the incorporation and ongoing administration of their companies and other structures, including accounting, tax returns and corporate secretarial services, and I serve as a professional director where required. With the support of a network of first-class cooperation partners, I also implement solutions worldwide.

My approach is to act as a trusted advisor who works directly with decision-makers, on an equal footing: I start from what you want to achieve and shape every structure, contract and tax decision around your commercial goals and how it improves your after-tax return. As “lawyers who make it happen”, my team and I stay involved until your company is set up, your structure is in place and your transaction has closed, and we remain available for any follow-on work that may be required. You benefit from bespoke solutions and personal attention, and sound, friendly and practical advice. We are not the lowest-cost provider in the market, nor do we intend to be; our aim is to deliver well-founded, proper solutions and excellent value for money.

Get in touch →