The organised community in the age of artificial intelligence: a pressenza reading

On 17 June 2026, the international press agency Pressenza published a Spanish-language essay by Argentine community organiser Jorge Nuñez Arzuaga titled La comunidad organizada en la era de la inteligencia artificial. In it, Nuñez Arzuaga argues that artificial intelligence is not a peripheral concern for movement-building but its new operating environment — one that is already redistributing power inside the neighbourhoods, factories and tenant associations of Argentina and the wider Global South. The piece is short, polemical and, read against the year's dominant AI coverage, quietly subversive: it treats AI as a political fact that organised communities must shape, rather than a productivity story that governments and vendors will settle between them.

The essay's wager is straightforward. For most of the past three years, mainstream discussion of AI has been conducted in three registers — the laboratory, the boardroom and the national-security apparatus. Pressenza, a cooperative news agency founded in Buenos Aires in 2009 and now publishing in Spanish, English, Portuguese, German and Italian, has consistently refused those frames in favour of a fourth: the meeting hall, the occupied factory and the soup-kitchen committee. Nuñez Arzuaga's intervention continues that line and gives it sharper technical edges, insisting that community organisations can no longer treat AI as someone else's problem.

A new tool, an old fight

Nuñez Arzuaga frames the current moment as the third rupture in modern movement infrastructure. The first, he writes, was the telephone and the mimeograph — tools that let neighbourhood and union organisers coordinate across a city at speed. The second was the social-media platform of the 2010s, which collapsed geography but introduced a dependence on private intermediaries whose commercial incentives were never aligned with the movements using them. The third, the essay argues, is generative AI: a layer that promises to translate, summarise, draft and analyse at marginal cost, but that also concentrates the means of production of language and image inside a handful of well-capitalised firms headquartered in the United States and China.

The argument is not Luddite. Nuñez Arzuaga accepts that the new tools are genuinely useful — for drafting leaflets in five languages, transcribing long community meetings, monitoring local council minutes, building simple case-management software for tenant unions. His complaint is that the conversations about how those tools will be deployed are happening almost entirely in rooms to which organised communities have no access. The result is a slow transfer of practical knowledge — how to write, how to record, how to find a case in a stack of files — from people who have spent decades doing it to systems whose behaviour cannot be audited by the people affected.

Counter-narrative: productivity, not politics

The dominant counter-frame, advanced through the year's policy papers and industry filings, treats AI primarily as a productivity shock. In this reading, the relevant decisions are technical — model architecture, compute allocation, safety benchmarks, export controls — and the relevant institutions are firms, frontier labs and a small number of finance ministries. Communities enter the picture, if at all, as downstream consumers of services or as labour markets to be re-skilled.

Pressenza's essay pushes back on that account in two ways. First, it observes that the productivity framing is itself a political artefact — one that benefits the firms selling the technology and the governments competing to host them. Second, it points out that the most consequential AI deployments of the past two years — algorithmic scheduling in warehouses, automated tenant-screening, predictive policing pilots — have not been the spectacular generative products that dominate press conferences. They have been the unglamorous, often invisible systems that decide who gets hired, who gets housed and who gets watched. Those are precisely the systems that organised communities have the standing and the technical knowledge to contest, if they can find the resources to do so.

Structural frame: who sets the defaults

The deeper argument in the essay, stripped of its communitarian vocabulary, is about default settings. AI systems encode assumptions about what counts as a legitimate goal, who counts as a relevant actor and what kind of evidence is admissible. Once those defaults are embedded in widely used tools, changing them requires technical and organisational capacity that most community organisations do not currently have. The essay warns that without deliberate investment in local AI capacity — open-source models, community-owned data cooperatives, multilingual training corpora drawn from the regions where the tools will be used — the defaults will be set elsewhere.

This is, in effect, a sovereignty argument: not the sovereignty of nation-states, but the operational sovereignty of the small collectivities that actually run meetings, distribute food and resolve disputes. Pressenza's wider editorial line — visible across its coverage of cooperativism, ecology and nonviolence — has long insisted that those collectivities are the legitimate units of democratic life. The new essay extends that position into the technical domain, contending that the question "who trains the model" is now as consequential for community organisers as the question "who owns the factory" was for the labour movement of the early twentieth century.

Stakes: a window that may not stay open

The practical stakes are concrete. In Argentina, where Nuñez Arzuaga is based, a wave of community-managed clinics, recuperated factories and neighbourhood assemblies has, over the past two decades, built a parallel infrastructure of mutual aid. If those organisations can deploy AI tools on their own terms — multilingual, locally hosted, auditable by the people using them — the technology could amplify an existing civic capacity. If they cannot, the same tools will be used against them: by landlords screening tenants, by employers scheduling shifts, by police departments predicting protest routes.

The essay's most provocative claim is that the window for the first outcome is narrow. Frontier-model development is consolidating around a handful of firms; open-source alternatives, while improving, are increasingly dependent on those same firms for compute and base weights. Without coordinated pressure from organised communities, the essay argues, the technology's trajectory will be locked in long before ordinary users understand what has been decided for them.

What remains uncertain

The Pressenza essay is, by its own admission, a provocation rather than a policy document. It does not name specific models, quantify deployment levels in Argentine community organisations, or compare alternative open-source stacks on cost or capability. The implicit empirical claim — that organised communities can, with modest resources, train and operate useful AI tools for their own purposes — is plausible but under-tested, and the piece offers no case studies to anchor it. Readers should treat it as a frame for further reporting rather than a verified finding.

What the essay does establish, clearly and in plain language, is that the political work of AI governance cannot be outsourced to firms or to finance ministries, and that community organisations have a legitimate standing in the conversation. Whether that standing translates into operational capacity is the open question for the rest of the decade.

Desk note: Pressenza is a cooperative news agency with roots in the Argentine peace movement; this piece treats its essay as primary source material for a debate that mainstream coverage has so far framed in purely technical or commercial terms. Monexus flagged it for the alternative frame — community organising as the unit of AI accountability — rather than for any specific factual claim that would need separate corroboration.

Wire provenance

This editorial synthesis draws on the following public wire/social posts:

• https://en.wikipedia.org/wiki/Pressenza
• https://en.wikipedia.org/wiki/Recuperated_factory
• https://en.wikipedia.org/wiki/Movement_building