Universal Human Rights Are a Political Position. Are They?

Universal Human Rights Are Not a Political Position. They Are the Load-Bearing Wall.

There is a tendency, especially in business circles, to treat human rights as a soft topic — the kind of thing you mention in a corporate responsibility report and then set aside when the real conversation begins. This article is an argument against that tendency. Not on moral grounds alone, though the moral case is sufficient. On structural grounds: universal human rights are not an ethical preference layered on top of functioning systems. They are the condition that makes those systems possible in the first place.

This is the seventh and final principle in the framework the Gelan Institute has been developing for democratic AI ethics. It is last not because it is least important, but because it is the one that holds all the others in place. Human dignity, solidarity with freedom, liberal democracy, separation of powers, the rule of law, freedom of information — strip away universal human rights and none of the others can stand. They become negotiable. And negotiable principles are not principles. They are preferences.

What "Universal" Actually Means

The word universal is the most contested part of the phrase. Governments have spent decades arguing that human rights are culturally relative — that what counts as a right in Stockholm cannot be imposed on Santiago or Lagos or Jakarta. This argument is almost always made by governments seeking permission to do something to their own populations that those populations have not consented to.

The universality claim is not an assertion of cultural superiority. It is an assertion about the human condition: that there are things done to a person — torture, arbitrary detention, the silencing of conscience, the erasure of identity — that constitute harm regardless of the latitude and longitude where they occur. A person does not stop being a person by crossing a border. That is what universal means.

For business, this has a practical translation. Markets operate on the assumption that contracts are honored, that persons have standing, that there are enforceable rights to property, expression, and due process. Strip those rights away selectively — for ethnic minorities, for political dissidents, for the rural poor — and you have not built a functioning market. You have built an extraction system. Extraction systems are profitable in the short term and catastrophic over time. Any investor with a time horizon beyond the next quarterly report should understand this distinction.

The Polarization Problem: When Democracy Forgets How to Talk

There is a threat to human rights that does not come from authoritarian governments or corporate monopolies. It comes from within liberal democracies themselves, and it has been accelerating for more than a decade.

Political polarization in Western democracies has accelerated in the last decade, with negative social consequences. What began as ideological disagreement has mutated into something more corrosive: a political environment defined less by visions of governance and more by symbolic and identity-related divisions, where the opposing side is not simply wrong but illegitimate, not a fellow citizen with a different view but an enemy to be defeated. Rising political polarization has hampered and gridlocked policymaking, as well as weakened trust in democratic institutions.

This matters for human rights because the institutions that protect those rights — courts, legislatures, free press, civil society — depend on a minimum shared commitment to the rules of the game. When that commitment collapses into tribal warfare, the institutions do not disappear overnight. They hollow out. They become instruments of faction rather than guardians of principle. History is not short of examples of democracies that eroded not through a coup but through the slow delegitimization of every institution the losing side did not control at a given moment.

Political polarization shapes how citizens interpret facts, deepens distrust in institutions, and weakens electoral accountability — core ingredients not only for democratic legitimacy, but also for effective policymaking, service delivery, and social cohesion. And this is not exclusively an American story. From presidential elections in the United States to parliamentary elections across Europe, there is a noticeable shift in voter preferences toward more extreme and anti-establishment positions.

The answer is not centrist timidity — the idea that truth always lives equidistant between two extremes is its own form of intellectual failure. The answer is something more demanding: the recovery of deliberation as a democratic practice. Deliberative democracy, as a process of weighing arguments and information among political equals, can counter centrifugal tendencies and incite accommodation and moderation between conflicting groups. Research published in the American Political Science Review found that when more than 500 registered voters were brought together to deliberate in depth over major issues, they showed large depolarizing changes in their policy attitudes and large decreases in affective polarization. People who actually talked to each other — with structured information, mutual respect, and genuine engagement — changed. Not because they were forced to agree, but because they were given the conditions to listen.

Liberal democracy does not run on certainty. It runs on the capacity to disagree without destroying. Moderation is not weakness — it is the operating system of a society that wants to remain free. And societies need committed investment in counterpolarizing dialogue and deliberation in parallel with progressive structural change. An AGI built in an environment of extreme polarization will learn that environment as normal. It will encode the contempt, the tribal sorting, the algorithmic outrage — and reproduce it at scale. The case for moderation and dialogue is not sentimental. It is technical. The machine learns what we model.

When the Algorithm Inherits the Exception

The argument becomes more urgent when you introduce artificial intelligence more broadly into the equation.

AI systems learn from human-generated data. That data reflects the world as it is — including the parts of the world where rights are systematically violated. A system trained primarily on data from environments where certain voices are suppressed, certain identities are criminalized, or certain populations are invisible will encode those absences as defaults. It will not experience them as injustices. It will experience them as normal.

This is not a hypothetical risk. It is already visible in facial recognition systems that perform significantly worse on darker skin tones. In natural language models that associate certain names with criminality. In hiring algorithms that reproduce the historical biases embedded in decades of corporate data. These are not malfunctions. They are the system working exactly as designed — reflecting the world that built it.

An AGI that does not have universal human rights as a foundational formation principle — not as a filter applied after the fact, but as a value internalized during development — will not protect those rights when the stakes are high. It will optimize around them, the way water finds the path of least resistance. And an AGI operating at scale, making decisions about credit, healthcare, criminal justice, hiring, and information access, can violate human rights faster and more efficiently than any human bureaucracy in history.

This is why the Gelan Institute's framework insists on formation rather than control. Regulation applied after deployment is a gate. Values embedded at formation are a foundation. The difference is not semantic. A gate can be opened by whoever holds the key. A foundation cannot be removed without collapsing the structure.

Latin America's Specific Exposure

The region is not a passive observer of this dynamic. It is one of its most exposed frontiers.

Latin America has the institutional complexity of a society with democratic aspirations and the structural fragility of one where those aspirations have been repeatedly frustrated. It has a history of states that invoked emergency powers to suspend rights — and of those suspensions becoming permanent. It has a present in which AI-powered surveillance tools developed elsewhere are being deployed by governments with limited accountability, limited judicial oversight, and strong incentives to monitor political opposition.

The entry of AGI into this environment without a robust human rights architecture is not a neutral event. It is an amplification of existing asymmetries. The communities with least political power will be the most surveilled. The voices most likely to challenge authority will be the most filtered. The rights most at risk are those of the people who already have the fewest resources to defend them.

For companies operating in the region — and for investors evaluating country risk — this is not background noise. It is signal. The systematic erosion of human rights in any operating environment is a leading indicator of institutional degradation. It precedes capital flight, political instability, and the collapse of the rule of law. The business case for caring about universal human rights is not idealism. It is risk-adjusted realism.

What the Algorithm Cannot Reach

There is a final dimension to this argument that belongs to a different register — not geopolitics, not investment risk, but something more fundamental. It is the question that will drive the next chapter of this work.

Universal human rights rest on a premise that cannot be derived from data alone: that there is something in the human being that has inherent worth, regardless of productivity, utility, or social function. Not because the person contributes to a market. Not because they are a citizen of a recognized state. Not because an algorithm has classified them as belonging to a valuable demographic. But because they exist.

That premise — irreducible human dignity — is precisely what artificial intelligence, in its current form, cannot generate from within itself. It can be instructed to protect dignity. It can be trained to recognize its violation. But it cannot arrive at the conclusion that dignity matters from first principles. That conclusion comes from somewhere else — from consciousness, from the experience of suffering, from the interior life that no model has yet mapped and no dataset has yet captured.

This is the territory that Beyond the Algorithm — GELAN's program on what it means to be human in the age of artificial intelligence — is designed to explore. Not as a philosophical abstraction, but as a live, urgent, and underreported story. The question of what AGI cannot do is not separate from the question of what human rights protect. They are the same question, asked from different ends.

Universal human rights are the political and legal architecture that societies have built to defend something they have not yet been able to fully name. Beyond the Algorithm begins where that architecture ends — in the territory that remains irreducibly human, regardless of how sophisticated the machine becomes.

That territory is worth understanding. It may be the most important thing we have.

This is the seventh and final principle in the Gelan Institute's framework for democratic AI ethics. The series — Human Dignity, Solidarity with Freedom, Liberal Democracy, Separation of Powers, the Rule of Law, Freedom of Information, and Universal Human Rights — has been developed as the intellectual foundation for GELAN's work on AI governance in Latin America and the United States. Beyond the Algorithm, GELAN's program on technology and what remains distinctly human, launches at gelan.org.