Ashnir Grover warns that introducing a fee for UPI could lead to the disappearance of mobile payments

The application of Artificial Intelligence (AI) in the traffic sector has evolved considerably on roads and urban streets in recent years. With the introduction of systems capable of examining images in real time, traffic authorities have gained greater capacity to identify potential infractions and risky behaviors.

However, this technological expansion has raised a frequent question among drivers: does artificial intelligence itself have the power to automatically issue fines? The established answer is negative.

Despite the technology presenting great accuracy and speed in detecting abnormalities, the issuance of infraction notices still depends on the evaluation and confirmation, either in person or remotely, by a traffic agent. In this context, AI functions merely as a support and screening tool for inspection, without replacing the function of human authority.

Modern monitoring systems use high-definition cameras combined with computer vision algorithms. These software programs are trained to recognize behavioral patterns on the roads and continuously analyze images to signal possible deviations from the Brazilian Traffic Code (CTB).

The operation of this process allows for the monitoring of thousands of vehicles simultaneously, increasing the effectiveness of inspection and enabling coverage of stretches where constant presence of inspectors would be impractical.

However, it is crucial to emphasize that image processing-based systems have technical limitations. Conditions such as poor lighting, reflections on the windshield, objects inside the vehicle, or the angle of the occupants can lead the algorithm to interpretation errors.

The check performed by the traffic agent constitutes the essential phase to discard these so-called 'false positives', ensuring that any penalty is applied only upon unequivocal proof of the infraction.

In Brazil, the formalization of the infraction notice is an administrative procedure directly linked to the traffic authority or its accredited agents, as determined by the CTB. Since AI does not possess legal personality or legal competence, it lacks the autonomy to impose sanctions directly and automatically on drivers.

The need for human review aims to ensure the legal security of the process and prevent system failures from resulting in unjustified charges. Furthermore, since not all traffic infractions can be detected by cameras, the technology is more efficient for visual infractions, while actions requiring physical tests remain the exclusive responsibility of field enforcement.