How Artificial Intelligence is Reshaping Power, Privacy, and Democracy
The world is not just. That is not a controversial statement. It is a historical fact etched into the lives of billions of people across the Global South. And now, just as humanity seemed to be slowly, painfully inching toward greater equality, a new force is emerging that threatens to lock existing inequalities into place permanently. That force is artificial intelligence.
In a candid conversation published on YouTube, Senator Bernie Sanders challenged Claude, Anthropic’s AI system, on the staggering scale of data collection happening invisibly in Americans’ daily lives. The response was stark. Companies are harvesting browsing histories, locations, purchasing habits, and even how long a person pauses on a webpage, feeding all of it into AI systems that build extraordinarily detailed personal profiles, largely without meaningful consent. “Most people click agree on terms of service without reading them,” Claude acknowledged to Sanders. The goal? Profit. Advertisers, political campaigns, and data brokers buy and sell this information to manipulate consumer behavior and, more alarmingly, voter behavior.
But this is not merely an American problem. It is a global one. And for the developing world, it is far more dangerous.
The pattern is devastatingly familiar. Every major technological revolution in modern history followed the same script. The Industrial Revolution did not liberate humanity. It powered the colonial machine that extracted wealth from Asia, Africa, and Latin America into European coffers. The internet promised democratization and delivered surveillance capitalism, controlled by a handful of corporations based in California and Beijing. Now AI is following the same trajectory, only faster, more invisibly, and with far greater reach.
Consider what is actually happening. Billions of people across developing nations, in Bangladesh, Nigeria, Indonesia, across the Global South, generate enormous amounts of data every day. Their browsing habits, their purchases, their social connections, their political opinions, all harvested continuously. That data flows into AI systems owned by American and Chinese technology corporations, training models that generate hundreds of billions of dollars in profit. Those profits do not flow back to Dhaka or Lagos. They flow to Silicon Valley and Shanghai.
This is data colonialism. It mirrors the historical extraction of raw materials from colonies with striking precision. The cotton grown in Bengal enriched Manchester. The data generated in Bengal today enriches Mountain View. The mechanism has changed. The relationship of power has not.
What makes this moment uniquely dangerous is that developing nations are not merely being exploited economically. They are being rendered technologically dependent in ways that undermine their sovereignty entirely. When a nation’s financial infrastructure runs on foreign platforms, when its government systems depend on foreign servers, when its young people are educated on foreign algorithms that quietly shape their values and worldview, that nation has lost something more fundamental than money. It has lost the capacity for genuine self-determination.
Historically, colonialism required physical presence. Armies. Administrators. Visible flags. The new colonialism requires none of that. It requires only a smartphone and an internet connection. It is colonialism without friction, without visibility, and therefore without the resistance that visible occupation historically provoked.
Meanwhile, the United States and China are engaged in fierce competition that has nothing to do with the welfare of ordinary people in smaller nations. Both offer partnerships and development assistance built on AI infrastructure. Both are really securing influence, data dependency, and geopolitical control. Countries like Bangladesh are caught between two imperial ambitions, with no genuinely independent option available.
Senator Sanders pushed Claude on whether a moratorium on new AI data centers might create the leverage needed to force genuine protections. Claude initially suggested targeted regulation as a more balanced approach. Sanders delivered a sobering reality check. AI companies are pouring hundreds of millions of dollars into the political process precisely to ensure those safeguards never materialize. Claude conceded the point directly. “You’re absolutely right, Senator. I was being naive about the political reality.”
That admission matters deeply. The companies building the most powerful AI systems are motivated by profit, accountable primarily to shareholders, and largely indifferent to questions of global justice. The poor subsidize the powerful. The many enrich the few. And it is happening invisibly, rebranded as progress and innovation.
Privacy, as Claude told Sanders, is not merely a personal issue. It is a democracy issue. And beyond democracy, it is a sovereignty issue for every nation that lacks the resources to build and control its own technological infrastructure. When companies and governments hold detailed profiles of millions of people, they hold power over those people in ways most citizens do not fully grasp. They can manipulate choices, predict behavior, and influence thinking at a scale that makes traditional propaganda look primitive.
The question is whether resistance is possible. Regional cooperation among developing nations to build shared technological infrastructure, demands for genuine data sovereignty, support for open source AI that no single nation controls, these are real possibilities. But they require political leadership that is genuinely independent of both Washington and Beijing. That is extraordinarily difficult to build in a world where most governments in the developing world are already deeply entangled in the competing orbits of the two great powers.
History shows that unjust systems, however powerful they appear, eventually face resistance. Empires fall. But they do not fall on their own. They fall when enough people clearly see what is being done to them and refuse to accept it as inevitable.
The first step is naming it honestly. What is being built right now, in the name of progress and innovation, is a new architecture of inequality more deeply embedded and harder to escape than anything that came before. Senator Sanders is asking the right questions in Washington. The rest of the world needs to start asking them too, before the answers are decided for us.
The world is not just. And without conscious resistance, AI will make it less so.
H M Arafat Ullah
He is a writer, researcher and public intellectual based in Dhaka, Bangladesh.
