The EPFL AI Program is an interdisciplinary initiative at the École Polytechnique Fédérale de Lausanne (EPFL), a Swiss public research university in Lausanne, Switzerland. Founded in 1969, EPFL is part of the Swiss Federal Institutes of Technology Domain, alongside its sister institution ETH Zurich, and operates under the Swiss Federal Department of Economic Affairs, Education and Research. The AI Program coordinates research, teaching, and innovation in Artificial intelligence, spanning fields such as Machine learning, Deep learning, and Neural network architectures, while leveraging EPFL's broader strengths in engineering, computer science, and life sciences.
The program builds on EPFL's established reputation in technology and science. As of 2025, EPFL enrolled 14,072 students from over 130 countries, and its faculty and alumni include a Nobel Prize laureate, two Fields Medalists, two Pritzker Architecture Prize winners, and one Turing Award winner. The AI Program is situated within this highly international environment, with French and English as official languages, and benefits from the university's urban campus extending alongside Lake Geneva, including the EPFL Innovation Park and affiliated research centers.
History and Evolution
The roots of EPFL trace back to 1853, when a private school named École spéciale de Lausanne was founded by Louis Rivier and John Gay. Over the following decades, it evolved through several reorganizations, becoming the technical department of the Académie de Lausanne in 1869, then the École d'ingénieurs de l'Université de Lausanne in 1890, and later the École polytechnique de l'Université de Lausanne (EPUL) in 1946. In 1969, EPUL separated from the University of Lausanne to become a federal institute under its current name, EPFL, directly controlled by the Swiss Federal Council.
The AI Program emerged from EPFL's expansion into modern computational fields. Under the presidency of Patrick Aebischer, who took office in 2000, EPFL began developing life sciences and related technologies, absorbing the Swiss Institute for Experimental Cancer Research (ISREC) in 2008. This period saw rapid growth in research areas that now underpin the AI Program, including Machine learning and Deep learning. The program reflects EPFL's broader trajectory: from 360 students in 1946 to 1,400 students and 55 faculty members in 1969, and over 13,000 full-time students as of 2023.
Research Areas and Initiatives
The EPFL AI Program encompasses a wide range of research topics, from foundational algorithms to applied systems. Core areas include Machine learning techniques, Deep learning architectures, and Neural network models, with particular emphasis on Large language model development and Generative AI applications. Researchers at EPFL contribute to advancing Transformer (architecture) models, which have become central to modern AI, and explore innovations in Multi-Head Attention, Positional Encoding, and Encoder-Decoder Architecture frameworks.
The program also investigates practical aspects of AI deployment, such as Model Pruning, Data Augmentation, and Learning Rate Scheduling optimization. It collaborates with industry partners and research institutions, including those in the broader Swiss and European ecosystem. Notable faculty and researchers associated with EPFL's AI efforts include experts in Machine learning and Deep learning, though the program's exact staffing and project portfolio evolve continuously. The AI Program leverages EPFL's strong rankings in engineering and technology, which have been recognized by QS World University Rankings and Times Higher Education.
Education and Training
The AI Program offers educational pathways at both undergraduate and graduate levels. Undergraduate students in computer science and related disciplines encounter AI fundamentals, including Machine learning and Neural network concepts. Graduate programs provide advanced training in specialized areas such as Deep learning, Large language model design, and Generative AI. Courses are taught in French and English, reflecting EPFL's bilingual environment.
Admission to EPFL follows the university's general policies. Swiss holders of a "maturité gymnasiale" are directly accepted into bachelor's programs, while international students must have a final grade average of 80% or above and demonstrate French proficiency at level B2 on the CEFR scale. The first year includes a propaedeutic cycle with a block examination; approximately 60% of students fail this first year across all majors, with higher failure rates in fields like Life Sciences Engineering, Physics, Mathematics, and Electrical Engineering. The AI Program benefits from EPFL's emphasis on entrepreneurship, with the EPFL Innovation Park supporting start-ups; since 1997, an average of 12 start-ups per year have been created by students and faculty.
Campus and Innovation Ecosystem
The AI Program operates within EPFL's campus at Dorigny in Écublens, construction of which began in 1974 and was inaugurated in 1978. The campus has expanded over the decades, with the first development stage completed in 1984 at a budget of 462 million Swiss francs. The EPFL Innovation Park, an integral part of the campus, fosters collaboration between academia and industry, providing a fertile ground for AI research and commercialization.
The program's location in Lausanne, on the shores of Lake Geneva, positions it within a vibrant European technology hub. EPFL's international outlook, ranked as the most international university in the world by Times Higher Education in 2014 and 2015, attracts talent from over 130 countries. The AI Program leverages this diversity to address global challenges in Artificial intelligence, from ethical considerations to practical applications in healthcare, robotics, and natural language processing.
Impact and Future Directions
The EPFL AI Program contributes to the university's strong performance in global rankings. In 2023, QS World University Rankings placed EPFL 16th worldwide across all fields and among the top 10 in several engineering disciplines. Times Higher Education ranked EPFL 19th globally for Engineering and Technology. The CWTS Leiden Ranking, which measures scientific impact, ranked EPFL 13th worldwide and 1st in Europe in 2013. The Times 100 Under 50 Rankings, which includes EPFL due to its 1969 founding, ranked it 1st in the world in 2015, 2016, and 2017, and 2nd in 2018 and 2019.
Looking ahead, the AI Program aims to advance the frontiers of Artificial intelligence through interdisciplinary research and collaboration. It continues to develop new methodologies in Deep learning, Large language model efficiency, and Generative AI systems, while addressing societal implications. As of 2025, the program remains a dynamic component of EPFL's mission to educate future leaders and drive innovation in science and technology.