First measures of the data science strategy implemented

The Federal Administration has drawn up the first measures of its federal data science strategy (DSStB): a report on concrete application cases and a code of practice for human-centric and trustworthy data science. The Federal Council acknowledged this at its meeting on 8 November 2023.

Six working groups with specialists from the Federal Administration and the ETH Domain have developed concrete application cases for data science in the Federal Administration. The application cases show the great potential of data science applications in the Confederation’s various policy and activity areas. Among other things, this involves the use of open source intelligence, holistic model systems for decision support and quality assurance of data through plausibility checks.

It highlights in a comprehensible and practical way how leaders from politics and the administration can use data science in their everyday work. In addition, the necessary framework conditions to apply these data at a broad and interdepartmental level are in place. Administrative units are thus supported in recognising the potential of data science for themselves and promoting its use.

Code of practice

A code of practice raises awareness among the Federal Administration’s administrative units of the core principles of human-centric and trustworthy data science. This is achieved by means of practical explanations and the implementation of these principles in everyday work. These core principles as defined in the data science strategy are: data and information protection, information security, data security, data governance, non-discrimination, explainability, traceability, transparency, reproducibility, neutrality, objectivity and ethical handling of data and results.

The code of practice also strives to increase the credibility of the administrative units’ work and strengthen public trust in the application of data science and specific methods from the field of artificial intelligence (AI). The code of practice is a voluntary code and not a binding set of instructions or legal regulations.

Great potential of data science

Numerous offices within the Federal Administration already use data science applications, for example, to calculate trends in passenger transport or to improve thunderstorm forecasts. The potential, however, is still great. The Federal Council therefore wishes to advance the use of data science in order to promote the public good and to support public policy.

With the data science strategy and the creation of both the Data Science Competence Center (DSCC) and the Competence Network for Artificial Intelligence (CNAI), over the past three years the Federal Council has laid important foundations across its departments for the human-centric, responsible and competent use of data science and AI in the Federal Administration.