To combat illegal dumping of trash, several local authorities and companies are turning to artificial intelligence. Initial trials show improvements, although the legal framework for these systems remains unclear.
The phenomenon concerns a significant portion of the population. According to the latest report from Gestes Propres, dated September 2025, an association fighting against littering, 35% of French people have discarded trash on the ground in the previous twelve months. For municipalities, these practices can quickly become a major problem, particularly due to illegal dumpsites, which harm safety and biodiversity.
To identify and penalize those responsible, more and more cities are installing AI-powered cameras. These systems can recognize objects and detect unusual movements, such as a car stopping for its passengers to throw out cardboard boxes or deposit an old washing machine. An alert can then be sent to law enforcement, who review the surveillance footage and, if applicable, record the offenders’ license plates. The person responsible for the dumping then receives a fine and has about ten days to contest the facts.
Fighting illegal dumping
The town of Gex, in the Ain department, has been testing this solution since 2024. According to reported information, it has reduced illegal dumping by 80% at sites where the smart cameras were installed. The system has led to 310 proceedings being initiated and 140,000 euros collected, with the average fine being around 500 euros.
In Orléans, a trial has been underway since the end of 2025. Seventy people have been penalized there, with fines that can reach up to 4,000 euros.
Other actors have specialized in monitoring dumpsites in the wild. This is the case for Disaitek, which analyzes satellite images to detect where dumpsites are forming and where they are becoming dangerous. This detection allows, following the same principle, for earlier intervention, particularly in the face of wildfires.
A usage framework still to be defined
The legal framework, however, remains uncertain. The Cnil, the data protection authority, reminds us that in France, the law does not authorize AI-enhanced cameras to detect offenses in public spaces.
Innovation can then come at another stage, once the waste has been collected. The startup Lixo proposes analyzing this waste using cameras installed in garbage trucks, which several dozen local authorities are equipped with. Each time a bin is emptied, photos are taken and an algorithm identifies non-compliant waste.
Data collected by the company shows that nearly one in two bins contains at least one anomaly. This could be, for example, a nitrous oxide canister, which can explode when compressed and endanger workers. The stated goal is not to penalize, but to enable the deployment of prevention campaigns.
