AI Camera Allegedly Misidentifies Dutch Motorist as Using Mobile Phone, Issuing €380 Fine

A Dutch smart camera on the A2 equipped with AI falsely identified a motorist as using a mobile phone while driving, issuing a €380 fine. The driver, Tim Hanssen, was scratching his head when the AI system misclassified the action.

Risk classification

• Primary risk domain: 7 AI system safety, failures, & limitations
• Primary risk subdomain: 7.3 Lack of capability or robustness

The AI system failed to perform robustly under real-world conditions, misclassifying a common human gesture (scratching head) as a traffic violation due to limitations in its training data.

Additional risk subdomains

• 5.1 Overreliance and unsafe use: The human police officer overrelied on the AI's classification and validated the incorrect fine without proper critical review.

Causal factors

• Entity: AI
• Intent: Unintentional
• Timing: Post-deployment

The incident was triggered by a false positive from the deployed Monocam AI system, which misidentified a head scratch as phone use.

EU AI Act risk tier

• Risk tier: 2 High Risk

High Risk: The system is used in law enforcement and traffic management, which are explicitly categorized as high-risk applications under the EU AI Act.

AI system and alleged parties

• AI system: None named
• AI purpose: Smart Surveillance; Identification
• Behaviour type: Tool
• Alleged developer: Unknown traffic enforcement camera technology developer
• Alleged deployer: Dutch police
• Alleged harmed parties: Tim Hanssen

Harm severity

Severity ratings are based on this scale.

Highest direct severity in any category: Negligible. Severity is scored from Negligible to Catastrophic in each harm category, for harm the reports describe as caused directly or indirectly by the AI system.

• Physical: direct Negligible, indirect Negligible
• Infrastructure: direct Negligible, indirect Negligible
• Property: direct Negligible, indirect Negligible
• Financial: direct Negligible, indirect Negligible
• Environmental: direct Negligible, indirect Negligible
• Malicious content: direct Negligible, indirect Negligible
• Differential treatment: direct Negligible, indirect Negligible
• Civil rights: direct Negligible, indirect Negligible
• Democracy: direct Negligible, indirect Negligible
• Privacy: direct Negligible, indirect Negligible
• Psychological: direct Negligible, indirect Negligible
• Epistemic: direct Negligible, indirect Negligible
• Child sexual exploitation and abuse: direct Negligible, indirect Negligible

People affected

• Occurrences reported: 1
• People reportedly harmed: 1
• People reportedly exposed: 1

Potential causes

Management

• Premature System Deployment: Deploying traffic AI before refining gesture recognition accuracy.
• Inadequate Reviewer Training: Reviewers were not trained to spot common AI gesture false positives.

Technology

• Algorithmic False Positives: The model falsely predicted phone use based on hand-to-ear gesture.
• Lack of Gesture Differentiation: Algorithm could not distinguish scratching head from holding a phone.
• Missing Confidence Indicators: System did not present model confidence scores to the human reviewer.

Data Inputs

• Biased Training Dataset: Dataset lacked images of drivers with empty hands near their ears.
• Overrepresented Phone Images: Training data had mostly phone-to-ear images causing pattern bias.

Human Factors

• Reviewer Automation Bias: The officer trusted the AI flag and failed to inspect the photo closely.
• Visual Misinterpretation: A dark shadow in the hand was mistaken for a phone during quick review.

Process and Methods

• Inadequate Verification Process: The manual review process failed to filter out the false positive.
• Inefficient Appeal System: Contesting wrongful fines takes up to 26 weeks to resolve.

Regulatory Environment

• Lack of Accuracy Standards: No regulations exist to mandate minimum accuracy for traffic AI.

Information quality

• Classification confidence: High
• Reason for confidence: The reports provide consistent, detailed accounts of the incident, including the specific fine amount, the driver's professional background, and a clear technical explanation of the false positive.

A Dutch driver was incorrectly fined after an AI-powered traffic camera misidentified a head scratch as mobile phone use. The incident highlights minor issues with automation bias and public trust in automated law enforcement, but poses negligible threats to national security.

• Overall national security impact: Minor
• Scope: Single nation
• Primary target: Netherlands
• Alleged perpetrator: Unknown

Threat characteristics

• Imminence: Long-term. Represents an ongoing technical limitation in deployed automated systems rather than an active national security crisis.
• Autonomy: Human-supervised. The AI system flags potential violations, but a human police officer is tasked with reviewing and validating the final citation.
• Novelty: Established threat. Computer vision false positives and human automation bias are well-documented, established technical issues.

Impact by dimension

• Physical security: Negligible. The incident involves a traffic camera misidentifying a hand gesture as phone use, with no physical threat or impact on critical infrastructure.
• Information security: Negligible. No intelligence compromise, classified data theft, or information warfare operations occurred during this incident.
• Sovereignty: Negligible. While representing an administrative error in local law enforcement, it does not threaten state authority, territorial control, or core government functions.
• Economic security: Negligible. The economic impact is limited to a single administrative fine of 389 euros, posing no threat to national economic stability or strategic industries.
• Societal stability: Minor. Indicates minor public trust issues regarding automated surveillance and law enforcement systems, but the issue is manageable within standard legal appeal procedures.