Security

No whining, just perform. For many athletes, that's a lesson they learn early on at their sports club. Standards like that travel with them, from youth teams to the Olympic podium. But they also have a downside: an increased risk of physical, psychological and sexual misconduct. The Winter Games show a glorious result. But how safe is the road to get there?

It’s almost 2026. As we prepare to ring in the new year, think back to the turn of the millennium in 2000. Remember? The world was gripped by the ‘millennium bug’. Computers were expected to go haywire, stock markets to crash, planes to fall from the sky. In the end, it wasn’t so bad. What people often forget is that this was thanks to billions spent on preparation.

Smart cities promise to make us safer. Cameras on every corner that detect violence and algorithms that alert the police before anyone calls. For many, that sounds reassuring. But how safe do these systems really make us, and more importantly, for whom?

Entering a drug lab without knowing which gases are being released is life-threatening. “That risk needs to be reduced,” says University of Twente researcher Dimitar Rangelov. Together with Saxion University of Applied Sciences, the Police Academy and the Technical University of Sofia, he is developing an AI-driven forensic investigation system that makes crime scene work faster, safer and more accurate.

AI helps clinicians quickly find the information they need, filtering hundreds of reports and documents, automating processes, and preventing something from being overlooked. But traditional AI isn't safe enough when it comes to high-level tasks, such as diagnosing. Here, explainability is crucial. Unlike non-explainable models that operate like black boxes, explainable AI follows a clinically acceptable reasoning and reveals how it arrives at a decision, so doctors can evaluate whether its output is trustworthy.