In recent years, debates on artificial intelligence have largely revolved around regulation, competitiveness and existential risks. Yet beyond the corridors of Brussels or Silicon Valley, the question of AI touches on something far more fundamental: whether innovation can serve the many rather than the few.
A striking case comes not from the usual technology hubs but from rural Colombia, where experiments with AI in education highlight both the promises and perils of digital transformation. These experiments invite us to revisit an underused ethical compass—omnipartiality, the demand that every perspective, particularly those at the margins, must shape collective decision-making.
This essay explores how AI in Colombian rural schools reveals the stakes of responsible innovation. By situating the case within broader global experiences and Europe’s own regulatory ambitions, I argue that omnipartiality should guide not only classroom technologies in the Global South but also Europe’s effort to align digitalisation with democracy.
The idea of omnipartiality has deep conceptual roots. John Rawls, in A Theory of Justice, famously asked us to imagine making decisions behind a “veil of ignorance”, stripped of knowledge about our position in society. While Rawls framed this as impartiality, omnipartiality goes further—it requires that every standpoint, including those most disadvantaged, be actively represented.
Jürgen Habermas added another dimension with his discourse ethics. For him, legitimacy arises only when all who are affected can participate in rational dialogue. The principle of inclusivity is not merely moral but procedural, shaping the very process of decision-making. Yet Habermas, like Rawls, often presumes a level of access and capacity that many communities, especially in the Global South, lack.
Iris Marion Young pushes us further still. In Inclusion and Democracy, she insists that marginalised voices must not only be present but must transform the terms of debate. For Young, inclusion is not tokenism but a recognition that perspectives born of exclusion hold epistemic value—they see what dominant groups cannot. Omnipartiality, then, is not neutrality but solidarity, an effort to reframe innovation from the standpoint of those most likely to be forgotten.
Colombia’s Experiment: Hope and Unease
In Colombia, where rural communities often struggle with underfunded schools, AI tools have been piloted to support teachers and students. On the surface, these projects seem promising. Adaptive learning platforms can tailor instruction to individual needs. AI-assisted marking can lighten the burden on overstretched teachers. Language models can help bridge gaps where teachers are scarce.
Yet the Colombian case also underscores serious risks. Many rural areas face weak internet infrastructure, leaving AI tools unreliable or inaccessible. Teachers report feeling sidelined when algorithms dictate learning paths without considering local context. More troubling still, students’ data are collected without robust safeguards, raising questions about surveillance and consent in communities already vulnerable to state and corporate overreach.
Omnipartiality reframes this dilemma. Instead of asking whether AI “works” in a narrow efficiency sense, the question becomes: whose needs and voices shaped its design and implementation? Did rural teachers help shape the curriculum algorithms? Were parents informed about data use? Did students themselves have any say in how technology would enter their classrooms?
Too often, the answer is no.
Colombia’s challenges resonate globally. In Kenya, AI-powered platforms have been introduced to rural schools with similar promises of personalisation. Yet studies reveal that when electricity is unreliable and teacher training limited, such tools can deepen inequalities rather than bridge them. Children in urban centres thrive whilst rural pupils fall further behind. Omnipartiality would demand that these infrastructural realities be treated as central, not peripheral, to the innovation process.
In Eastern Europe, pilot programmes in countries like Romania and Bulgaria have faced another challenge—cultural mismatch. AI platforms designed in English-speaking contexts often fail to reflect local languages and pedagogical traditions. Teachers, rather than being empowered, feel alienated. Here again, responsible innovation requires not simply importing technology but co-designing it with educators and students on the ground.
These cases demonstrate that AI in education is never neutral. It can reproduce inequalities as easily as it can reduce them. What links them is the absence, or presence, of omnipartial thinking. When the most vulnerable are included from the outset, technology bends toward justice. When they are ignored, innovation becomes yet another instrument of exclusion.
Why should Europe care about Colombian classrooms or Kenyan experiments? For one, Europe positions itself as a global leader in responsible AI. The EU AI Act, recently finalised, sets out the world’s most ambitious attempt to regulate algorithmic systems. Its emphasis on transparency, risk assessment and human oversight rightly addresses many dangers posed by unaccountable AI.
Yet regulation is only part of the story. Europe also invests heavily in digital education through the Digital Education Action Plan and seeks to align technology with the broader goals of the Green Deal and social inclusion. If Europe is serious about exporting a model of responsible innovation, it must learn from contexts where the stakes are highest—places where technology can either widen or narrow the gap between privilege and marginalisation.
Colombia, Kenya and Eastern Europe remind us that responsible innovation cannot be confined to elite policy debates. It is tested in classrooms with patchy internet, among teachers juggling multiple jobs, and in communities where trust in institutions is fragile. Europe’s credibility as a normative power in AI will depend not just on the stringency of its regulations but on its willingness to engage with and learn from these realities.
Building Omnipartiality Into Innovation
What would it mean to embed omnipartiality in responsible innovation? Three principles stand out.
First, co-design from the margins. Innovation must begin not with developers in corporate headquarters but with those most affected. For rural schools, this means teachers, parents and students should co-shape the goals and design of AI systems. In practical terms, this could involve participatory workshops, local pilot committees and feedback loops that have real power to alter projects.
Second, context over universality. Too often, AI tools are exported as if they were universally applicable. Omnipartiality demands attention to context—infrastructure, culture, language and history. A system that works in Barcelona may not work in rural Colombia, and vice versa. Responsible innovation must embrace diversity, not erase it.
Third, ethics as ongoing practice, not checkbox. Omnipartiality resists the idea that ethics can be “built in” once and for all. Instead, it treats inclusion as an ongoing practice of questioning, adjusting and expanding who gets to shape technology. For Europe, this means creating mechanisms for continuous dialogue with affected communities, both within and beyond its borders.
The story of AI in Colombian rural schools is not simply about technology—it is about democracy. When innovation is imposed without the voices of those most affected, it echoes older patterns of domination, whether colonial, corporate or bureaucratic. When it is shaped omnipartially, it becomes a site of empowerment, where marginalised communities do not just receive technology but actively redefine its purpose.
For Europe, the lesson is clear. If the promise of responsible innovation is to be more than rhetoric, omnipartiality must become central. This means that debates on AI should not only include experts, regulators and industry leaders but also those who live at the edges of power—rural teachers in Colombia, children in Kenyan villages, educators in small towns across Eastern Europe, and indeed marginalised communities within Europe itself.
Omnipartiality challenges us to imagine innovation differently. It shifts the question from “What can technology do?” to “For whom and by whom is technology made?” The Colombian case demonstrates that without this shift, even well-intentioned projects can reproduce inequality. Global comparisons show that this is not an isolated problem but a structural one. Europe’s own ambitions demand that it takes these lessons seriously.
Responsible innovation will not be measured by glossy strategy papers or regulatory milestones alone. It will be judged by whether technology expands or narrows the circle of democratic inclusion. Omnipartiality offers the compass to ensure the former. It is time Europe, and indeed the world, took it seriously.
Deny Giovanno is a public affairs strategist and political thinker based in Jakarta. He writes on ethical leadership, Southeast Asia’s democratic traditions, and the legacy of Sutan Sjahrir and Indonesia’s early socialists. His work draws on experience in both corporate advocacy and civil society engagement.
