Algorithms are playing an increasing role in Swiss public administrations, particularly in healthcare, justice, insurance and school dispatch. A report from AlgorithmWatch details these uses, including AI to diagnose cancer, chatbots for administrative requests, and preventive tools for security.
In Switzerland, algorithms are used in a variety of public sectors to support decision-making, from crime prevention and medical diagnosis to insurance management and the allocation of students to schools. The AlgorithmWatch organization recently published a report on the uses of algorithmic decision support systems (ADM) in the Swiss public sector.
In the healthcare sector, for example, Swiss hospitals such as HUG and Nyon are using artificial intelligence (AI) to diagnose cancer and select appropriate treatments, with tools such as IBM's Watson for Genomics. In Zurich, ADM is also used in radiology and pathology.
Chatbots are also deployed in cantons such as St. Gallen and Aargau, facilitating the management of administrative requests, particularly those related to insurance premium reductions. This type of tool is designed to lighten the workload of public services faced with a large number of requests.
In the criminal justice sector, FaST software helps prevent recidivism by classifying offenders according to their risk. The data used includes general statistical factors such as criminal history, age and other risk indicators.
Some cantonal police forces, such as those in Basel-Landschaft, Aargau and Zurich, use preventive software to identify areas at risk of burglary. The PRECOBS program, for example, analyzes burglary trends to predict possible future targets.
ADM systems are also used by the Federal Customs Administration to analyze the risks associated with goods, and determine which should be inspected. This process will be further enhanced by the DaziT program, designed to digitize all customs procedures by 2026.
In the insurance sector, notably for accident and military insurance, algorithms could soon automate decision-making by evaluating medical records and calculating premiums.
Last but not least, projects such as the one on the distribution of pupils in Zurich schools show how ADM can be used to promote social diversity within schools, by adjusting the composition of classes using data analysis algorithms.
Applications of algorithms in public administrations in Switzerland are multiplying, raising questions about the ethics, transparency and accountability of decisions made by automated systems.
Source : ICTjournal
