to what extent will we accept that AI cameras fine us

Spain has begun to automate the surveillance of minor infractions with AI:

Technology stops optimizing traffic and starts monitoring it.

Why is it important. This marks a paradigm shift. Until now, urban AI was used to improve mobility (adjust traffic lights, predict traffic jams, reduce emissions).

Now he goes from assistant to inspector. And it does so with a key nuance: it does not pursue major crimes or flagrant dangers, but rather small daily infractions that previously escaped control due to cost and surveillance capacity.

AI reduces the marginal cost of sanctioning to practically zero. Once deployed, everyone can be observed all the time.

The facts. The Barcelona pilot test involved four buses of lines H12 and D20 equipped with cameras that identify, through AI, vehicles blocking reserved lanes. In Madrid, the City Council has installed smart traffic lights that count pedestrians in real time and has announced systems that will detect seat belt use.

The DGT has taken another step. It has deployed four cameras on the A-1, A-2, A-6 and A-42 highways that monitor the crossing of continuous lines. The system works with two cameras per section: one records the license plates at the beginning, another at the end. If a car changes lanes between both points, the fine is automatic. It is 200 euros per violation.

In figures. Spain already has 3,395 devices to control violationsaccording to Faconauto.

• Of them, more than 1,300 are DGT surveillance points between fixed and mobile radars.
• Added to this are more than 200 cameras that monitor belts and mobile phones, Pegasus helicopters and now these new continuous line detection systems.

Barcelona has not yet activated the sanctions on its buses, but the volume of violations detected (80 daily in just four vehicles) anticipates what is coming.

Between the lines. There is a delicate balance that is being renegotiated without us having barely opened the debate.

1. On the one hand, more compliance with fewer agents: administrative efficiency is indisputable.
2. On the other, the sensation of an omnipresent eye.

The difference with the classic radar is not so much technical as range.

• The radar monitors specific points where there is proven risk.
• These new systems turn the entire city into a guarded zone: each bus is an inspector, each intersection a control point.
• AI does not change what is sanctioned, it changes where and how much.

Move from selective surveillance to ubiquitous surveillance.

Yes, but. To what extent will citizens accept being recorded and punished by a machine? It’s not just a legal issue, but a cultural one: trust in the algorithm versus human interpretation. Who audits the system’s decisions? What room is there for error or appeal?

Technology is not neutral: each deployment reflects political priorities about what deserves to be monitored and sanctioned.

The big question. What is relevant is not whether this is good or bad in the abstract, but what it tells us about the new contract between citizen, city and AI.

• AI stops being an abstraction and enters the daily urban experience.
• The citizen goes from user to observed subject.

And the unresolved question is who sees, who decides, who corrects and, above all, how far we are willing to go when automating the chase is so easy and cheap.

In Xataka | The “made in China” business of the DGT’s V-16 beacons: homologating the same product 24 times and selling it under different brands

Featured image | Barcelona City Council