Data School and Research Software Lab awarded grant for ‘Responsible AI in Practice’
A research team from Data School and the Research Software Lab, both part of the Centre for Digital Humanities (CDH), has been awarded a €125,000 grant from the SIDN Fund and Topsector ICT for their project Functioneringsgesprek voor AI (FG-AI).
Organisations are increasingly expected to demonstrate that their AI systems perform well and are reliable. The FG-AI project – which translates as Performance Review for AI – introduces a method to help organisations better monitor their AI systems. It functions much like a traditional performance review, but for algorithms and AI. Through regular evaluations, organisations can assess whether the AI continues to operate as intended over time. Does the algorithm still serves the organisation’s goals? Does it still align within current legal and social standards?
The research team has already developed a first prototype. With the new funding, they will focus on academic publications, a best practices report, and developing an open-source web application. To this end, a long-term pilot will be launched with eight to twelve public-sector partners, each of whom will monitor at least one algorithm over the course of a year.
Project partners
The Data School and the Research Software Lab are collaborating in this project with consultancy firm Berenschot and the regulatory authorities Netherlands Food and Consumer Product Safety Authority (NVWA) and the Dutch Authority for Digital Infrastructure (RDI). The project team includes Iris Muis, Julia Straatman, Bart Kamphorst, and Mirko Schäfer (Data School); Jill Briggeman, Sheean Spoel, and Mees van Stiphout (Research Software Lab); and Lisa de Graaf (Berenschot).
Responsible AI in Practice
In total, the SIDN Fund and Topsector ICT have awarded €1.1 million in funding to ten project proposals as part of their Responsible AI in Practice programme.