Barrancabermeja adds AI drones to its urban surveillance, but the announcement doesn't say what that artificial intelligence does

By Zendoric · July 31, 2026.

The mayor of Barrancabermeja, Jonathan Vásquez Gómez, announced the addition of two drones with artificial intelligence (AI) to the security system of the Colombian oil district. As the mayor himself explained, the equipment was arranged through the Colombian Federation of Municipalities and will reinforce the District Monitoring Center, currently supported by video surveillance cameras and panic buttons installed in various areas.

The Secretariat of Security and Coexistence is coordinating the initiative, which —according to local authorities— seeks to optimize surveillance of public spaces, track risk situations and improve response capacity to incidents affecting public order. The announcement comes alongside a reinforcement of National Police patrols in the city's comuna 3, as the police force itself reported.

What stands out, and this is our reading, is what the announcement does NOT say. It does not specify exactly what the AI on these drones does: whether it is facial recognition, automatic detection of objects or people, movement-pattern analytics, or simply video transmission with automatic tagging. Nor does it mention the technology provider, the cost of the operation or, above all, the legal framework that will govern the handling of data captured on citizens who have committed no crime.

That opacity is not a minor detail: it is the pattern. In general, the rollout of algorithmic surveillance in mid-sized Latin American cities is moving faster than the data protection rules that should accompany it. Political announcements tend to present drones and cameras as synonymous with modernization and security, but they rarely include independent audits, data retention limits or accountability mechanisms for how the footage is used.

This connects to a tension we have flagged in other analyses: AI applied to public safety can be a legitimate tool —helping to detect fires, accidents or crimes in real time and save lives— or it can become a control infrastructure deployed without citizens knowing what is being watched, who accesses that data and for what purpose. The difference between the two scenarios is not set by the technology itself, but by the governance surrounding it, and in this case that governance appears nowhere in the announcement.

In the short term, cases like this are an honest reminder that AI adoption by local governments often advances without the necessary institutional checks: transparency about the algorithms, judicial oversight of facial recognition use, or clear limits on storing images of ordinary people. In the long term, tools of this kind —properly audited and with clear rules— can be part of safer, more efficient cities, consistent with the promise of an AI that frees up resources and human time instead of concentrating them in opaque hands. But that horizon only arrives if the discussion about the rules of the game moves at the same pace as the purchase of the drones, and today, in Barrancabermeja, there is no sign that this is happening.