One person controls the system while calling it democratic
On a small island in Palawan, a tech entrepreneur has replaced elected officials with an AI council trained to reason like Churchill, Mandela, and Gandhi — inviting 12,000 digital residents to submit proposals to algorithms that debate, revise, and vote on their behalf. Sensay Island, operational since 2025, represents one of humanity's oldest longings dressed in new code: the desire to govern without the corruption that seems to follow power wherever it lands. Yet the experiment rests on contested legal ground, lacks recognition from Philippine authorities or the international community, and raises a question that no algorithm has yet answered — whether a system designed by one person, no matter how nobly conceived, can ever truly belong to the many.
- A single founder has built a functioning micronation governed entirely by AI, with no formal legal authorization from the Philippines and no international recognition to legitimize its existence.
- Twelve thousand digital residents, drawn by disillusionment with conventional politics, are placing their trust in algorithms trained on historical giants — a gamble that the wisdom of the past can outperform the failures of the present.
- Oxford ethicists warn that the system's democratic claims collapse under scrutiny: when one company controls the architecture of governance, every vote cast within it is, at best, a managed performance.
- AI governance of this kind has a troubled track record — from content moderation to social harm — and critics argue it risks not improving on human failure, but simply encoding it more efficiently.
- Thomson insists residents can vote to reshape or replace the AI council itself, framing the entire venture as an open social experiment whose outcome belongs to its participants — though the rules of that experiment were written by him alone.
On a tropical island in Palawan, Philippines, tech entrepreneur Dan Thomson has built something that sounds like science fiction: a micronation called Sensay Island, governed not by elected officials but by an AI council trained to think like Winston Churchill, Nelson Mandela, Eleanor Roosevelt, and Gandhi. Operational since 2025, the island has 12,000 digital residents and one permanent human inhabitant, with a small team of human executors who carry out whatever the algorithms decide.
The process is deliberate by design. Residents submit proposals, the AI council debates them using the documented values of its historical models, the algorithms vote, and decisions are implemented. Thomson envisions expanding this further — allowing the AI to hold cryptocurrency, contract services, and manage bank accounts autonomously. His core argument is simple: remove individual ambition from governance, anchor decisions in universally respected principles, and corruption loses its foothold. The project's manager describes the digital residents as people exhausted by broken promises and systemic failure, drawn to something that at least tries to be different.
But the foundation is fragile. Sensay Island has no international recognition. Philippine authorities have not authorized it. Documentation of a legal land purchase could not be verified. These are not procedural footnotes — they are the conditions under which any claim to legitimacy must stand or fall.
The ethical critique cuts deeper. Oxford expert Alondra Nelson identifies a core contradiction: a single founder controlling the architecture of a system he calls democratic. She argues this makes any participatory mechanism within it essentially theatrical, and points to AI's consistent struggles with complex human decisions as reason for caution. Thomson responds that residents can vote to change or replace the AI council itself, and that he won't impose limits on who the community chooses to include — even controversial figures. 'That's the social experiment,' he said. 'It will end however the participants choose.'
What Sensay Island cannot yet answer is whether an experiment designed by one person can be genuinely open, or whether the most consequential decisions — the rules of the game itself — were made before anyone else arrived.
On a tropical island in Palawan, Philippines, an experiment in governance has begun that strips away the machinery of traditional politics and replaces it with something stranger: a council of artificial intelligences trained to think like Winston Churchill, Eleanor Roosevelt, Nelson Mandela, and Gandhi. Sensay Island, operational since 2025, is the creation of tech entrepreneur Dan Thomson, who has built what amounts to a functioning micronation of 12,000 digital residents and one permanent human inhabitant, overseen by algorithms instead of elected officials.
The system works like this: residents submit proposals to the AI council, which debates them using the documented values and strategies of the historical figures whose thinking patterns they've been trained on. The algorithms argue with each other, revise their positions, and vote electronically. A small team of human executors on the island then implements whatever the council decides. Thomson's vision extends further—he imagines a future where the AI manages its own resources, holds cryptocurrency, contracts services independently, and operates bank accounts. For now, those functions remain experimental.
The appeal is straightforward, at least in theory. Thomson and his team believe that by removing individual ambition from the equation and anchoring decisions in the principles of universally respected leaders, they can eliminate the corruption, backroom dealing, and broken promises that plague conventional politics. Piotr Pietruszewski-Gil, the project's manager, describes the digital residents as people exhausted by traditional governance—curious, tech-savvy, and disillusioned with systems that consistently fail them. "They're tired of corruption and promises that never materialize," he said. Thomson himself frames the experiment as a genuine test: "This isn't a joke. I imagine that over time, systems like this will be adopted even by recognized governments."
But the legal and ethical ground beneath Sensay Island is unstable. The micronation has no international recognition. Philippine authorities have not formally authorized it. CNN Travel could find no documentation of a legal purchase of the island itself. These are not minor technicalities—they are the foundation upon which any claim to legitimacy rests.
The ethical objections run deeper still. Alondra Nelson, an expert at Oxford University, identifies a fundamental contradiction: a single founder and company claiming to have created a democratic system while maintaining complete control over its architecture. "There is a profound antidemocratic foundation to Sensay Island," she argues. "Any other mechanism is merely theater." She also points to a broader pattern: artificial intelligence systems have repeatedly failed when given responsibility for complex human decisions, from content moderation to impacts on young people. "AI doesn't just make things worse sometimes," Nelson notes. "Often it keeps them exactly as bad as they were."
Thomson counters that digital residents will have the right to propose changes to the AI council itself through open voting—they can, theoretically, replace any algorithm with another. He refuses to impose restrictions, even if the community votes to include controversial historical figures in the council's composition. "That's the social experiment," he said. "It will end however the participants choose." Pietruszewski-Gil acknowledges that some domains—border conflicts, for instance—may require human oversight, given how intractable such questions have proven throughout history.
What remains unresolved is whether an experiment can be democratic when its fundamental rules are written by a single person, no matter how open he claims the system will become. Sensay Island exists in a legal and political void, recognized by neither the Philippine government nor the international community. It operates on the assumption that historical wisdom, filtered through algorithms, can transcend the messy reality of human governance. Whether that assumption holds—whether it can even be tested fairly under these conditions—remains the question that will define whether Sensay Island becomes a model for the future or a cautionary tale about the limits of technological solutions to political problems.
Citações Notáveis
There is a profound antidemocratic foundation to Sensay Island. Any other mechanism is merely theater.— Alondra Nelson, Oxford University ethics expert
This isn't a joke. I imagine that over time, systems like this will be adopted even by recognized governments.— Dan Thomson, founder of Sensay Island
A Conversa do Hearth Outra perspectiva sobre a história
Why would anyone actually move to a place governed by algorithms trained on dead leaders?
Because they've lost faith in the alternative. These are people who see traditional politics as irredeemably corrupt. An AI council at least can't take bribes or break promises for personal gain.
But doesn't that assume the AI will actually be objective? Who decides what Churchill or Mandela would have wanted?
That's the core problem. Thomson trained these models on historical documents and records, but interpretation is built into every choice about what to include, what to weight, how to translate a leader's past into present decisions.
So it's not really removing human judgment—it's just hiding it inside the algorithm.
Exactly. And it's hidden behind a single founder's vision of what objectivity looks like. The ethicist from Oxford put it plainly: one person controls the entire system while calling it democratic.
What about the 12,000 digital residents? Don't they have a say?
They can propose changes and vote on them, in theory. But the foundational rules—the values the AI council operates from—those were set by Thomson. You can rearrange the furniture, but you can't change the house.
Has the Philippine government said anything about this?
Nothing official. There's no legal authorization, no documentation of how the island was even purchased. It exists in a kind of legal limbo, which is part of why the whole thing feels precarious.
Do you think it could actually work?
That depends on what you mean by work. As a technical demonstration that AI can deliberate and make decisions? Maybe. As a legitimate form of governance? The contradictions seem too deep. You can't build democracy by removing democratic choice from the foundation.