In 2009, I founded Datacell in Reykjavik — a data center that would become internationally known for a reason I hadn't anticipated when I signed the lease. Among our clients was WikiLeaks, the whistleblowing organization that had begun publishing classified documents that would shake governments around the world.
What followed was one of the most significant legal battles for digital freedom in European history, and the lessons from that experience have only grown more relevant as we enter the age of artificial intelligence.
The Blockade
In late 2010, after WikiLeaks published a massive trove of U.S. diplomatic cables, Visa and Mastercard imposed a financial blockade on the organization. They instructed their payment processors worldwide to stop handling any transactions related to WikiLeaks. For Datacell, which processed donations to WikiLeaks through our Icelandic payment infrastructure, this meant our legitimate business operations were being shut down by private corporations acting without any court order or legal basis.
The blockade was unprecedented. Two private companies had effectively decided to cut off the financial lifeline of a media organization — not because of any criminal conviction, not because of any court order, but because of political pressure. It was, in essence, extrajudicial economic censorship.
The Court Battle
We fought back. Datacell filed suit against Valitor (formerly Visa Iceland) in Reykjavik District Court in June 2012. Our argument was straightforward: Valitor had broken its contractual obligations by unilaterally blocking our payment processing without legal justification.
In July 2012, the district court ruled in our favor. The judge found that Valitor had indeed violated contract law by imposing the blockade. But the story didn't end there — the case was appealed to Iceland's Supreme Court, which in April 2013 upheld the lower court's ruling. Valitor was ordered to resume processing payments within two weeks.
It was a landmark victory — not just for WikiLeaks or Datacell, but for the principle that private financial infrastructure cannot be weaponized to silence speech without due process.
Parallels to the AI Era
Fast forward to 2025, and the questions raised by the WikiLeaks case have multiplied. Today, a handful of companies control the cloud infrastructure, AI models, and payment systems that the global economy depends on. The power to deny service — to "deplatform" — has become one of the most consequential forms of power in the digital age.
Consider the implications for AI. When a small number of companies control access to the most powerful AI models, they effectively control who can innovate, who can compete, and who can participate in the AI economy. When cloud providers can terminate accounts at will, they hold the power to shut down businesses overnight. When payment processors can block transactions, they can starve organizations of resources without any judicial oversight.
The principle we fought for in that Reykjavik courtroom — that infrastructure providers cannot arbitrarily deny service without legal basis — is the same principle that must guide how we govern AI infrastructure.
Building Independent Infrastructure
This experience is part of why I'm so passionate about building distributed, independent infrastructure at AI Green Bytes. Concentration of infrastructure is concentration of power. When all your compute runs in one company's cloud, when all your payments flow through two card networks, when all your AI models come from three providers — you're not just accepting a technical architecture. You're accepting a power structure.
Edge data centers, distributed across multiple jurisdictions, powered by renewable energy, and operated independently — this isn't just a technical strategy. It's a statement about the kind of digital future we want to build. One where no single entity can flip a switch and shut you down.
The WikiLeaks case taught me that digital freedom isn't abstract. It's infrastructure. And infrastructure is something we can build.
