A Dutch motorist Tim Hansen was wrongly fined 380 euros by an AI-powered traffic camera system called Monocam that mistakenly identified him as using his phone while driving when he was actually just scratching his head.
In October 2023, Tim Hansen, a Dutch IT specialist who works with AI image analysis algorithms at the firm Nippur, was photographed by a police AI camera system called Monocam while driving on the A2 highway near Weert. The system flagged him for using a mobile phone while driving, resulting in a 380 euro fine plus 9 euros in administrative costs that he received in November 2023. Upon reviewing the photographic evidence from the Central Judicial Collection Agency, Hansen discovered he was not holding any device but was simply scratching the side of his head with his right hand. The Monocam system, which has been deployed by Dutch police since July 2021, uses artificial intelligence to detect drivers using mobile phones and has caught 116,000 drivers texting in 2022 alone, with numbers believed to be higher in 2023. The AI system made a false positive identification, and this error was compounded by human error when a police officer reviewing the photo failed to notice that no phone was actually present in Hansen's hand. Hansen, who is left-handed and has hands-free calling capability in his lease car, filed an appeal against the fine, which can take up to 26 weeks to process. He wrote a detailed blog post explaining how such AI algorithms work and what likely went wrong, suggesting the training dataset probably contained many examples of people with phones near their ears but few examples of people with empty hands near their ears.
Domain classification, causal taxonomy, severity scores, and national security assessments were LLM-classified and may contain errors.
AI systems that fail to perform reliably or effectively under varying conditions, exposing them to errors and failures that can have significant consequences, especially in critical applications or areas that require moral reasoning.
AI system
Due to a decision or action made by an AI system
Unintentional
Due to an unexpected outcome from pursuing a goal
Post-deployment
Occurring after the AI model has been trained and deployed