TMNL has policies in place and implements processes to control risks when using data analysis models. For example, in each model, TMNL seeks to use the minimum amount of data required and looks at the potential risks of discrimination. Banks also decide in their role as data controller whether the model is appropriate. In carrying out these analyses and making decisions, TMNL and banks are assisted by the model ethics committee.
What does the model ethics committee do?
The model ethics committee was formed to advise TMNL independently and objectively on ethical issues when using models. The committee allows TMNL to reflect thoroughly on its use of models, such as on questions about the use of analytical techniques or dilemmas in implementation.
The model ethics committee currently comprises four academics, each contributing to the discussions from their own field of expertise. This enables them to identify any ethical issues, offer multiple perspectives and arguments on these topics, and provide well-considered, objective advice to TMNL and banks.
The ultimate goal is to ensure responsible AI practices are adhered to when using models. This means that, in addition to detecting potentially unusual transactions, other interests must be taken into account, such as those of specific social groups or related to privacy. These must be balanced by taking a thorough look at the various interests involved. The four renowned experts provide a boost of confidence in the work that we do.
Jeroen van den Hoven
Jeroen is a professor of Ethics and Technology at Delft University of Technology. He is also Editor-in-Chief of Ethics and Information Technology, Scientific Director of the Delft Design for Values Institute, Founding Scientific Director of the 4TU.Centre for Ethics and Technology, and a permanent member of the European Group on Ethics (EGE) for the European Commission.
Hinda Haned is an Endowed Professor of Data Science at the University of Amsterdam and Scientific Co-Director of the Civic AI Lab. Her research focuses on developing solutions and best practices in secure and responsible data science. Hinda is also founder of Owls & Arrows, a consultancy firm specializing in data science and auditing models and algorithms.
Mara is an expert in anti-money laundering and countering terrorist financing (AML/CTF). Her doctoral research focused on countering terrorist-financing policies, public-private partnerships, and the risk-based approach to countering terrorist financing in Europe (2008-2013). More recently, she has conducted various research and consultancy projects for Dutch and European authorities. She currently works as a Senior Intelligence Analyst at Swedish fintech Acuminor and is a member of the Dutch National Risk Assessment (Nederlandse Nationaal Risk Assessment, NRA).
Joras Ferwarda is an Assistant Professor at the Utrecht University School of Economics. He is also a member of the academic board of the international Criminology PhD programme for the Catholic University of Milan. He was recently a Visiting Professor in the Department of Criminology and Criminal Justice at the University of Maryland, College Park and Senior Researcher in Criminology at the Vrije Universiteit Amsterdam. He is a certified money-laundering expert for the Dutch National Police.