Recent events have cast a shadow of doubt over the infallibility of artificial intelligence in traffic law enforcement within the Australian Capital Territory. Multiple instances have emerged where drivers were wrongly accused by the ACT government's AI-powered cameras, leading to unwarranted fines and demerit points. These cases underscore a growing concern regarding the precision and fairness of these automated systems, prompting calls for a thorough re-evaluation of their operational integrity and the human oversight involved.
Details of the Disputed Fines: AI's Recurring Errors
In a striking incident on May 28, 2026, Fiona Devers, a 66-year-old resident with an unblemished driving history, found herself at the receiving end of an infringement notice. While navigating her Toyota Corolla along Gungahlin Drive, she was flagged by an AI camera for alleged mobile phone use. The penalty? A substantial $548 fine and three demerit points. However, a crucial piece of evidence from the same system, a photograph, distinctly showed Ms. Devers with both hands firmly on the steering wheel. She asserts she does not even own a smartphone, relying on an outdated Nokia, which remained unused for months. Upon presenting her case to Access Canberra, three employees concurred that the evidence against her was unfounded. Prompt action by The Canberra Times, inquiring about her situation, led to the immediate withdrawal of her fine.
This is not an isolated occurrence. Just the previous week, The Canberra Times reported on another driver who was issued a similar fine on Gungahlin Drive. His explanation? He was simply changing gears in his manual vehicle, an action misinterpreted by the AI as operating a mobile phone. Despite his pleas, he described the response from Access Canberra as unsympathetic, with officials unable to specify what he was supposedly holding. Earlier in May, a woman from Gungahlin faced a $548 fine for "mobile phone use - hand-held," even though she relies on Bluetooth for hands-free communication. These local incidents mirror issues in Sydney, where camera systems inaccurately read number plates, leading to Canberra drivers being wrongly charged for toll roads they never used. Though Transurban, the motorway operator, acknowledged the errors as rare and apologized, a comprehensive explanation of the technical glitches and the full extent of affected drivers remains elusive.
These recurring misidentifications by AI traffic cameras signal a critical juncture for automated enforcement systems. While the intent is to enhance road safety, the repeated errors highlight an urgent need for greater transparency, improved technological accuracy, and a more robust human review process. The experiences of individuals like Ms. Devers serve as a powerful reminder that while technology offers efficiency, it must be rigorously scrutinized to prevent arbitrary penalties and ensure public trust in the justice system. Moving forward, authorities must prioritize not only the deployment of advanced technology but also the development of foolproof mechanisms to correct algorithmic biases and prevent wrongful accusations, thereby upholding fairness and accountability for all road users.