A license plate reader system used by Aurora Police Department misidentified a family's minivan as a stolen motorcycle, leading to officers drawing guns on and handcuffing children ages 6-17 in a traumatic encounter.
On August 2, 2020, Aurora Police Department officers in Colorado conducted a high-risk traffic stop on a Black family after their license plate reader system flagged their minivan as matching a stolen vehicle. The system had matched the license plate number to a stolen motorcycle from Montana, despite the vehicles being completely different types. The incident involved four children ages 6, 12, 14, and 17, along with adult driver Brittney Gilliam, who was taking the children to get their nails done. Officers approached with guns drawn and ordered everyone out of the vehicle, placing some children face-down on hot pavement and handcuffing several of them. The children could be heard crying and calling for their mother in video recorded by a bystander. Over a dozen officers eventually responded to the scene. After determining the vehicle was not stolen, officers immediately removed handcuffs, explained the mistake, and apologized. Police Chief Vanessa Wilson stated it was not a profiling incident but rather 'a hit that came through the system.' The confusion may have been partly due to the fact that Gilliam's car had been reported stolen earlier in the year but was recovered the next day. The department has offered to pay for therapy sessions for the traumatized children and opened an internal investigation.
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