Medical AI earns trust only when its legal and ethical preconditions are designed in from the start. Mauritz Kop taught exactly that module at the SmartHealth masterclass Kunstmatige intelligentie in de zorg (October 29–30, 2020, Van der Valk Houten) — a two-day program for the healthcare professionals and innovators who buy, build and validate clinical AI in Dutch hospitals.
Law as a design input for clinical AI
Privacy, IP, liability and fundamental rights: the masterclass treated legal requirements not as after-the-fact paperwork but as design inputs for AI applications that touch patient data and patient safety. A diagnostic-support algorithm is never just a model — it is a GDPR processing operation, frequently a medical device under the MDR or IVDR, and a node in a liability chain that runs from developer through hospital to treating physician. Reading a product through all of those lenses before procurement was the working method of the module.
An interdisciplinary faculty
Clinicians and builders from Pacmed, Radboudumc, Attendi and Syntho taught alongside the legal module — grounding artificial intelligence governance in the daily reality of Dutch hospitals and health-tech startups, from intensive-care prediction models to synthetic data as a privacy-preserving alternative to raw patient records. Legal teaching that never touches a confusion matrix stays abstract; technical teaching that never touches the lawful-basis question ships unusable products. This program put both in one room.
From White Paper to binding law
The 2020 timing mattered: the European Commission's AI White Paper had just framed the risk-based approach, and healthcare AI was widely expected to land in the high-risk tier. The EU has since adopted the AI Act, whose high-risk obligations — risk management, data governance, human oversight, conformity assessment — now interlock with the MDR/IVDR regime, turning the masterclass's design-it-in message from policy signal into binding law on a phased timetable.
A continuing teaching line
From Maasstad Hospital to SmartHealth: educating the people who buy, build and use medical AI is a structural part of AIRecht's practice — see also the Artificial Intelligence Impact Assessment, the ex-ante review framework that anticipated exactly this compliance-by-design logic. The lesson travels beyond healthcare: a system whose legal preconditions are an afterthought meets its regulator at the worst possible moment, while a system that carries its compliance evidence by design meets the same regulator with the file already open. For anyone deciding on medical AI procurement today, that difference is the whole game.
Meer lezen