On Wednesday 12 to stop coming up with new advice 2025 The Award for the Most Brilliant Failure in the field of AI in the public sector was presented for the fourth time during the iBestuur AI and Government conference in Utrecht.. This election is organized by Platform AI & Government in collaboration with the Institute for Brilliant Failures.
The importance of a learning approach
AI offers many opportunities, also within the government. It is important that we share knowledge and new technologies such as (generative) Using AI to better help citizens and companies. Only by working together and experimenting can we accelerate the responsible use of AI.
Creating awareness for the importance of learning and experimenting with AI is the reason to pay attention to AI projects in a positive way. By actively sharing experiences, we learn from each other in order to develop socially relevant AI applications more efficiently and effectively in a responsible manner.
After an explanation by Paul Iske, jury chairman and professor of Individual and Organizational Learning at Maastricht University and also our Chief Failure Officer, the prize was presented by Sabine Herbrink, program manager Digitally Thoughtful, employed at ECP.
Award Award 2025 to the 'Predictive Policing' project’
The prize for the most brilliant failure 2025 was won by Prof. Dr. Wim Hardyns, professor of Criminology at the Ghent University and Safety Sciences at the University of Antwerp and responsible for the BIGDATPOL research programme.
This pilot study examined for the first time the use of 'predictive policing' by a Belgian police zone. It showed the technical feasibility and predictive power of AI models, but at the same time made it clear that an accurate algorithm is not enough to have an impact in practice. Due to organizational limitations, capacity problems and reluctance among agents, the predicted added value of AI was not fully utilized.
Sharing experiences and insights
We congratulate Wim Hardyns congratulations on being awarded the Brilliant Failures Award 2025! A more detailed description of this case can be found in BriMis. We also congratulate the nominated Prof. Dr. André Dekker, Professor Clinical data science aan de University of Maastricht and connected to Maastro. His submission concerns the use of an AI algorithm to detect COVID-19, that worked in Wuhan, China, but not in the Netherlands. You will find the description of this case here as well as an overview of important insights from various projects from which we can learn a lot. https://www.briljantemislukkingen.nl/wp-admin/post.php?post=19104&action=edit#
“AI requires a learning government: grow together, fail together, moving forward together. Only by working together and experimenting can we scale AI applications responsibly and successfully”, thusSuzie Kewal, department head of the Digital Society Directorate the Ministry of the Interior and Kingdom Relations.
The Institute for Brilliant Failures and Platform AI & Government would like to thank all parties who submitted a case for their openness and sharing their experiences to help others learn from their approach.
Award 2026
AI platform & Government and the Institute of Brilliant Failures underline the importance of experimenting with AI and promote this (brilliant) failure in the responsible development and application of AI is inevitable. We therefore invite all parties in the public sector to compete for the Brilliant Failures Award for 2026. Registrations are open from now until 31 March 2026 via our website.
