In the DSI Insight column on inside-it.ch, DSI Professor Anastasiia Koloskova explores how machine learning can be applied where it is most urgently needed. In healthcare, environmental science or the financial sector, data is often sensitive, scattered and difficult to access. Algorithms offer a solution, enabling institutions to train AI models collaboratively without exchanging raw data. This research approach ensures that data protection, efficiency and performance can be guaranteed simultaneously.
One of our overarching goals is to contribute to the democratisation of AI: to make advanced machine learning methods accessible – not only to well-resourced environments, but also to public institutions, research laboratories and other organisations operating under real-world practical constraints. For me, this is not just a technical issue, but also a social one: who benefits from AI – and who gets to help shape it?
Anastasiia Koloskova
DSI and Assistant Professor of AI and Optimisation at the University of Zurich