Colón, Panama

There is an irony worth noticing. The city where, in 2026, young Afro-descendant people from Latin America and the Caribbean articulated a shared political agenda is called Colón. The port bearing the Spanish name of the navigator whose voyages inaugurated the colonial order is today a centre of Afro-Panamanian identity, also shaped by the Afro-Caribbean communities who built the transcontinental railway and the Canal. There, a regional youth movement returned to the question of how that inequality is inherited and how it can be turned into a political programme.

The 2026 Afro-Latin Caribbean Youth Congress was also held on the bicentenary of the Amphictyonic Congress of 1826. The first attempt to articulate the Americas politically returns, two centuries later, through the voices of those whom that first congress did not name.

What this youth movement is demanding is not simply to be seen. It is asking for a voice in decision-making, public policy and regional governance. And it does so at a moment when the international framework on the rights of people of African descent has moved into an unexpected terrain, artificial intelligence.

The fourth session of the Permanent Forum on People of African Descent was organised under the theme of reparatory justice in the age of artificial intelligence. The UN and its anti-racism mechanisms have warned that digital systems can reproduce systemic inequalities when they are designed, trained and regulated from historically unequal structures.

That framework lands very precisely in the Colón declaration. Among its calls to action is the need to promote ethical and regulatory frameworks in artificial intelligence to prevent algorithmic racial profiling, digital discrimination and technological violence that disproportionately affects Afro-descendant youth, especially young Afro-descendant women.

The anthropological question, then, is no longer only how a population is represented. It becomes how a population is read, sorted, filtered and governed by technical systems. Databases, credit scoring, surveillance, automated procedures, access to services, content moderation and classification systems inherit a history they claim not to have.

The warning does not come from nowhere. From critical data studies, danah boyd and Kate Crawford argued more than a decade ago that big data does not guarantee truth, objectivity or precision by itself. More data does not necessarily mean better data. Outside context, data loses meaning. And the fact that something is accessible does not automatically make it ethically legitimate.

The demand from Colón enters through that door. It is not asking for a more diverse AI as a cosmetic gesture, but for a form of governance capable of asking who classifies, with which data, under what history and with what consequences.

For decades, multicultural policies treated recognition as a problem of visibility. The task was to name, represent, include and celebrate. But being visible does not guarantee being recognised as a rights-bearing subject. One can also be visible as risk, profile, suspicious category, exploitable data or market segment. The social difference these systems inherit does not disappear when it is digitised. It becomes automated.

References for this section

Brazilian Amazon

The second scene is in the Brazilian Amazon, where the physical demarcation of the Kawahiva do Rio Pardo territory has begun. This is the home of an uncontacted Indigenous people. It looks like a classic case of political anthropology. Territory, state, extractive frontier, logging, cattle ranching, road, decree. Precisely for that reason, it deserves close attention.

Demarcation is not simply a line on a map. It is a technical operation that decides what exists legally.

Here an exact paradox appears. A people who avoid contact, issue no statements and do not sit down to negotiate need to be made legible by the very state from which they flee in order to remain uncontacted. It is James C. Scott’s idea of legibility turned upside down. Usually, the state measures and catalogues in order to govern. Here it must measure and catalogue in order not to govern, in order to withdraw a forest from an economy that advances by cutting it down.

Survival depends on files, coordinates, markers and administrative deadlines. Once the physical demarcation is completed, one legal step will still remain, the presidential decree formally creating the Indigenous territory.

Refusing contact is not the absence of history. It is an extreme way of defending a life against economies that advance by measuring, occupying and turning forest into resource. Non-relation can also be a political relation. And Indigenous territory is not merely the material support of a culture. It is its condition of possibility. Without forest there is no language, kinship, memory, hunting or mobility that can be sustained. Cultural recognition, when it comes too late, can become an epitaph.

References for this section

Indonesia

The third scene is in Indonesia. Asia Centre and International Media Support have documented how climate disinformation affects Indigenous peoples in specific ways. It is not the freshest hook in today’s radar, because the report was published in January, but it works as a background piece for understanding the same problem from Asia.

The mechanism is familiar, although it is often narrated in language that is too clean. National development is invoked, modernisation is promised, large projects are presented as inevitable and Indigenous practices are discredited as backwardness, inefficiency or an obstacle to progress.

The key point is not only that false information circulates. Certain narratives produce political reality. If a community appears again and again as an obstacle to development, its expulsion can be presented as rational management. If its environmental knowledge is described as superstition, losing the territory can look like necessary modernisation.

Here development should be read not only as public policy, but as a language of power. Arturo Escobar showed that the anthropology of development cannot simply evaluate whether projects work or fail. It must also ask how the discourse of development itself defines which lives appear as backward, which territories become available for intervention and which forms of knowledge are treated as obstacles. In Indonesia, climate disinformation works precisely on that boundary. It turns Indigenous vulnerability into resistance to progress.

This point connects directly with digital anthropology. The struggle over territory no longer takes place only in courts, assemblies, roads or forests. It also happens in media, platforms, technical reports, campaigns and recommendation systems.

The struggle over climate is increasingly a struggle over the conditions of credibility. Who can speak. What counts as evidence. Which knowledge is taken seriously. Which story becomes common sense.

References for this section

Common reading

The three scenes seem different. A youth declaration in Panama. A demarcation in Brazil. A report on disinformation in Indonesia. But together they say the same thing with precision. Recognition is no longer played out only on the cultural plane. It is decided through technical layers.

In Colón, that layer is the algorithm. In the Amazon, it is the state cadastre and the decree. In Indonesia, it is the information flow that distributes the roles of vulnerable and obstacle. In all three cases, a community is not merely disputing its image. It is disputing the infrastructures that make its existence legible, protectable or disposable.

This is the lesson for thinking artificial intelligence from anthropology. It is not enough to ask whether machines represent cultures well. The harder question is what technical systems do with the social differences they inherit. Do they correct them, hide them or automate them?

A risk algorithm, a demarcation map or a climate database does not simply describe the world. It orders it, makes it governable, sometimes protects it and sometimes condemns it. Technique does not arrive after already formed cultures. It participates in the production of the social.

This is partly what Nick Couldry and Ulises A. Mejias call data colonialism. Their proposal is not simply to say that platforms extract information. The argument is stronger. Social life becomes a continuous source of raw material for systems of accumulation, classification and prediction. The colonial analogy matters because it points to a continuity in the appropriation of other worlds through languages of progress, efficiency and innovation.

Colonialism does not return only through old symbols. It can also return as metric, infrastructure, platform, procedure and apparently neutral technical narrative.

That is why the answers cannot be merely cultural either. Afro-descendant youth are demanding participation in technological governance, not simply celebration. Indigenous peoples need effective territorial recognition, not patrimonial admiration. Communities harmed by climate disinformation need to dispute the regimes of truth that decide what counts as development.

Anthropology has something to contribute here, provided it does not settle for commenting from outside. It can show that there is no data without history, no territory without relations and no climate without politics. And it can remind us that the technical is never only technical when it classifies people, distributes risks and decides which lives deserve protection.

The final question is not whether societies will be recognised by technology. It is who designs the conditions of that recognition. Who names, who measures, who demarcates, who verifies, who decides whether a community appears as a subject of rights, a remnant of the past or an obstacle to the future.

In Colón, in the Brazilian Amazon and in Indonesia, what is at stake is not only being seen. It is existing in a way that cannot be reduced, exploited or erased by the infrastructures that claim to administer the world.

References for this section

Sources