Colombia’s National Institute of Legal Medicine and Forensic Sciences has confirmed that three of the ten people killed in an air strike in Guaviare department on 27 August were minors. Two were fifteen. One was sixteen.

Colombian outlets reporting the medical examiner’s findings have put the number of minors killed in air strikes since the beginning of August at around fifteen. The government has said 24 suspected fighters have been killed in operations since President Abelardo de la Espriella took office on 7 August.

Those two sets of figures overlap, and the overlap is the story that international coverage of the offensive has mostly skipped.

The recruitment figures underneath

Human Rights Watch published a report last week finding that at least 1,500 children have been recruited by armed groups in Colombia since 2021, with 625 cases recorded in 2024 alone. The organisation describes recruitment increasingly running through social media, and children being given progressively more dangerous tasks — including, it says, operating drones carrying explosives.

Put those findings next to the medical examiner’s, and the picture is not ambiguous. If armed groups in southern Colombia contain a substantial number of recruited children, then aerial bombardment of their camps will kill children. That is a foreseeable consequence rather than an accident, and it is the central legal and moral problem with the strategy.

What the law actually says

This is contested territory, and it is worth setting out the positions rather than asserting one.

Under international humanitarian law, a child recruited into an armed group and taking a direct part in hostilities can lose the protection civilians enjoy, for as long as that participation lasts. Recruiting them was a war crime committed by the group that did it. That does not automatically make them unlawful targets afterwards.

The Colombian government’s position, in substance, is that camps of armed fighters are military objectives, that the state did not put children in them, and that suspending operations wherever children may be present hands armed groups an incentive to recruit more of them as protection.

The counter-position, argued by Colombian rights organisations and by the country’s own child protection bodies, is that knowledge changes the calculation. Where a commander knows in advance that a camp is likely to contain recruited minors, the proportionality assessment required before a strike has to account for that, and an air strike is the method least capable of distinguishing between a commander and a fifteen-year-old.

Colombia has been here before. Bombings that killed recruited children brought down a defence minister in 2019 and forced a public reckoning over what the armed forces knew beforehand. The precedent is domestic and recent.

The wider policy turn

De la Espriella won June’s election by less than a percentage point on a promise to confront armed groups militarily and to end his predecessor’s negotiated approach, known as total peace. Since taking office he has authorised repeated air strikes, described a strike in the Amazon that killed 20 as the most extensive in years, and announced results on social media.

The International Crisis Group, in a briefing published as he took office, argued that Colombia does need a more effective military strategy but that hardline methods alone tend to produce cycles of bloodshed and abuses against residents, and that negotiations, however frustrating, have historically been the more reliable route to unwinding armed groups. Supporters of the offensive answer that four years of talks coincided with those groups reaching their greatest strength since the 2016 peace accord.

Both readings are available on the evidence. What is not in dispute is that violence against civilians has risen alongside the operations, with retaliatory attacks, kidnappings and forced confinement reported in Norte de Santander and Cauca, and a car bomb at a police station this week.

A regional pattern

The displacement logic here is familiar from elsewhere in the region. Pressure applied in one place moves armed groups rather than removing them — the same dynamic we described in Michoacán, where guarding one crop moved the extortion next door, and along the Bolivia-Brazil border, where Brazilian pressure appears to have pushed gang violence across.

The detail HRW raises about children operating explosive drones also belongs to a broader shift: armed groups across Latin America are adopting cheap uncrewed systems quickly, and the people sent to fly them are frequently the most expendable members of the organisation.

What to watch

Whether the forensic institute continues publishing the ages of the dead. It is a state body, its findings have repeatedly contradicted the government’s framing of these operations, and it is currently the only source establishing how many of the people counted as fighters were children.

That is a narrow thing to watch, and it is the one that determines whether anyone outside Colombia ever learns the number.

Sources

• Colombian National Institute of Legal Medicine and Forensic Sciences findings, reported 30–31 August 2026
• Human Rights Watch report on child recruitment in Colombia, August 2026
• International Crisis Group Briefing 56, “From Total Peace to Iron Fist”, 7 August 2026
• Al Jazeera, AFP and Colombia Reports coverage of the August operations