Medicine is fundamentally an information problem. Human biology generates data in a volume and complexity that no individual doctor can process: genome sequences, blood values, imaging, medical histories, drug interactions. AI can.
The opportunity for Switzerland lies in the fact that it brings together two worlds that are often separate elsewhere: cutting-edge medicine and cutting-edge computer science.
By 2045, AI will be better at diagnosing rare diseases than the best human specialist. Already today, AI outperforms radiologists in detecting certain tumours in imaging [1]. The machine's advantage: it forgets nothing, it never tires, and it can search the entire medical literature in seconds.
For rare diseases -- conditions affecting fewer than 1 in 2,000 people -- this is particularly relevant. Patients with rare diseases today wait an average of five to seven years for a correct diagnosis [2]. AI can drastically shorten this odyssey by matching symptom patterns across millions of cases.
Treatment plans based on the genetic profile, medical history and real-time data of the patient will become standard. Instead of "one-size-fits-all" treatments, every patient receives a tailored therapy. Costs decrease because misdiagnoses and unnecessary medications are eliminated.
The Swiss Personalized Health Network (SPHN), supported by the university hospitals and the Swiss Academy of Medical Sciences, is already laying the groundwork for this [3].
The first replacement organs grown from stem cells will be routinely used in Swiss top-tier medicine. Research is well advanced: in 2023, researchers succeeded for the first time in growing functional mini-organs (organoids) that serve as test environments for drugs [4]. The next step -- transplantable organs from the patient's own stem cells -- would eliminate rejection by the immune system and dissolve transplant waiting lists.
USZ has developed AI systems for clinical diagnosis together with ETH Zurich. Focus areas are imaging (radiology, pathology) and prediction of post-operative complications [5].
The Centre Hospitalier Universitaire Vaudois cooperates with EPFL in the areas of neuroscience and AI-powered psychiatry. The combination of clinical experience and technological innovation makes Lausanne a centre for medical AI research [6].
Novartis and Roche, two of the world's largest pharmaceutical companies, already use AI in drug development. AI accelerates the identification of promising molecules from years to months [7]. Switzerland as a pharmaceutical hub thus has a natural advantage: the infrastructure, knowledge and regulatory experience are already in place.
The greatest societal effect lies not in extending life, but in reducing healthcare costs.
Switzerland's healthcare expenditure amounts to around 90 billion francs per year -- over 11 per cent of GDP [8]. A significant portion of these costs arises from:
AI can reduce all three cost drivers. Early detection prevents expensive late-stage treatments. Automated administration saves working hours. Personalised therapy avoids wrong treatments.
The average life expectancy in Switzerland currently stands at 84 years [9]. The combination of AI diagnostics, personalised medicine and regenerative medicine (stem cell organs) could raise this to over 95 years -- not in care beds, but in vitality.
This has consequences:
The technology will drastically improve and reduce the cost of basic medical care. But the main beneficiaries will not automatically be patients -- rather the companies that own the medical AI systems. Without political regulation, a two-tier healthcare system looms: first-class AI diagnostics for those who can pay, standard treatment for the rest.
Switzerland, with its mandatory health insurance system (KVG), has a structure that can ensure AI advances benefit everyone -- if the political will is there [10].
[2] EURORDIS: The Voice of 12,000 Patients. Rare Disease Survey, 2009.
[3] Swiss Personalized Health Network (SPHN): Strategy 2024--2028.
[5] University Hospital Zurich / ETH Zurich: AI in Clinical Medicine Partnership.
[6] CHUV / EPFL: Centre for AI in Healthcare, Lausanne.
[7] Novartis: AI in Drug Discovery. Roche: Computational Sciences.
[8] Federal Statistical Office (FSO): Healthcare costs and financing, 2023.
[9] Federal Statistical Office (FSO): Life expectancy in Switzerland, 2024.