The former public servant had worked with artificial intelligence, so he knew how a machine could misread a shadow. When his son was fined for using a mobile phone while driving on Gungahlin Drive, he examined the images from the automated camera. The photograph showed the young man’s open hand moving from the gear shift to the steering wheel of a manual Kia Cerato. There was no phone. To prove it, the father recreated the scene: he photographed his son’s empty hand in the same position, the car seat webbing clearly visible, the grip action absent. It was, he argued, an optical illusion—a gear knob mistaken for a device. Access Canberra reviewed the evidence and dismissed it. The $548 fine and three demerit points stood. For six months, the family appealed, and each time a human reviewer looked at the machine’s output and saw a crime.

That stubbornness is not unique to Australia. In Mexico City, new traffic regulations now punish drivers who manipulate a phone or play music at high volume while the vehicle is moving, with fines reaching 35 times the daily minimum wage. The logic is straightforward: a moment of distraction can alter everything. Yet the same rigid enforcement that penalises a driver for holding a sushi roll or wearing a dark jumper that obscures a seatbelt also erodes the very trust it seeks to uphold. When a citizen can demonstrate that a camera captured a manual gear shift rather than a mobile phone, and the bureaucracy still refuses to withdraw the infringement, the system begins to look less like a safeguard and more like a trap.

Across the Pacific, a parallel debate is unfolding in Argentina’s classrooms. At least eleven provinces have introduced restrictions on mobile phones in schools, and the municipality of Córdoba now requires devices to be switched off and stored out of reach during the school day. The measure responds to a stark reality: 59 per cent of third-graders already own a phone, and the figure rises to 90 per cent in secondary school. A report by the NGO Argentinos por la Educación, drawing on international research, notes that such bans do reduce distractions and screen time in class. But the evidence on whether they improve learning is, in the report’s careful phrasing, “ambiguous.” Some studies find modest gains for the most vulnerable students; others detect no effect at all. The technology is removed, but the hoped-for transformation does not always follow.

What links these stories is a growing public impatience with institutional inflexibility, whether the decision-maker is an algorithm or a judge. In Canada, the case of Omar Abdul Singateh has become a flashpoint. Awaiting sentencing for beating and torturing a man in Montreal until he handed over cryptocurrency, Singateh was released on bail by a Quebec judge. While free, he was named in a U.S. indictment for allegedly smuggling 100 guns from Florida into Canada. Last weekend, he was arrested in Toronto after an exchange of gunfire and a carjacking. A retired Canadian judge called it “an embarrassing case.” In Manitoba, a 19-year-old was bound with duct tape, beaten and robbed; four suspects were arrested and then released on court orders. Viewed from Buenos Aires or Brisbane, the pattern is familiar: systems designed to protect the public are seen to fail, not because the rules are absent, but because their application lacks the human nuance that justice demands.

The father in Canberra keeps a photograph on his phone. It shows a hand resting on a gear shift, fingers relaxed, no object in sight. It is a mundane image, a record of an ordinary gesture. A machine looked at that same scene and saw a violation. A human reviewer, trained to trust the technology, confirmed it. The photograph remains, a quiet testament to the distance between what a camera captures and what a citizen knows to be true.