In February 2026, the Indian government issued new regulations requiring social media platforms to remove content deemed “unlawful” within three hours of receiving a government order. Three hours. Not three days or three weeks — three hours. The previous requirement had been thirty-six hours, which was already among the most aggressive content removal timelines in the world. Meta alone had restricted over 28,000 pieces of Indian content in the first half of 2025 following government requests.
In the same month, across the world, the European Union continued to enforce the Digital Services Act — a sweeping regulatory framework that requires platforms to remove “illegal content” promptly, to maintain detailed records of their content moderation decisions, and to submit to audits by EU regulators. The EU’s AI Act adds further layers of oversight for algorithmic systems that influence what people see and read. Together, these regulations give European bureaucrats unprecedented power over the flow of information across an entire continent.
Meanwhile, Elon Musk’s X (formerly Twitter) received 72,703 government requests to remove accounts in 2024. Under Musk’s ownership, X has positioned itself as a free speech platform, but the reality is more complex. A 2025 University of Edinburgh study found that censored accounts on X lost 25% of their engagement and 90% of their follower growth. The platform complied with 83% of censorship requests from non-liberal governments in its first year under Musk — up from 50% in 2022. Even the world’s most vocal free speech advocate finds it difficult to resist the combined pressure of governments that control market access.
This is a History Future Now article. So we are going to do what we always do: look at the present through the lens of the past, and then look at where the future is heading. The question is simple. What happens to civilisations that control speech? And what happens to those that do not?
The historical record is unambiguous. And it is not comfortable reading for anyone who believes that the state should decide what its citizens are allowed to say, hear, or think.
Part 1: Athens — The Power of the Open Agora
In the fifth century BC, a small city on the Aegean peninsula produced more intellectual, artistic, and political innovation in a single century than most civilisations produce in a millennium. Athens gave the world democracy, philosophy, theatre, history, rhetoric, geometry, medicine, and the foundations of Western science. The names alone tell the story: Socrates, Plato, Aristotle, Sophocles, Euripides, Thucydides, Hippocrates, Pericles.
Athens was not peaceful. It was not gentle. It was not tolerant by modern standards — it excluded women, slaves, and foreigners from political participation. But it had one quality that distinguished it from every other city-state in the ancient world: it allowed, and indeed celebrated, open public debate.
The agora — the central public space of Athens — was where citizens gathered to argue about politics, philosophy, ethics, and anything else that interested them. The theatre staged plays that openly mocked political leaders, questioned the gods, and explored the darkest aspects of human nature. The law courts operated on the principle that any citizen could bring a prosecution and any citizen could defend themselves through public argument. The philosophical schools — Plato’s Academy, Aristotle’s Lyceum — operated as open forums where ideas were tested through rigorous dialectical debate.
This culture of open argument was not accidental. It was structural. Athenian democracy required citizens to make decisions on matters of war, peace, taxation, and public works through direct vote in the Assembly. To vote intelligently, citizens needed information and argument. Restricting debate would have undermined the entire political system. The Athenians understood, intuitively, that the quality of decisions depends on the quality of the information available to decision-makers.
The results speak for themselves. Athens, with a population of perhaps 300,000 (including slaves and non-citizens), produced an intellectual and cultural output that shaped the entire subsequent trajectory of Western civilisation. No comparable city in the ancient world — not Sparta, not Corinth, not Thebes — came close.
Part 2: Sparta — The Cost of Control
Sparta provides the counter-example. The Spartans created perhaps the most controlled society in the ancient world. Every aspect of Spartan life was regulated by the state. Boys were taken from their families at age seven and subjected to a brutal military training regime (the agoge). Speech was controlled — the term “laconic” (from Laconia, the region around Sparta) literally means “using few words.” Foreigners were periodically expelled (xenelasia) to prevent contamination by outside ideas. Travel abroad was restricted. Innovation was discouraged. The Spartan constitution, attributed to the semi-mythical Lycurgus, was treated as sacred and essentially unchangeable.
Sparta produced formidable soldiers. It won the Peloponnesian War against Athens. But it produced virtually nothing else. No Spartan philosopher, poet, playwright, historian, scientist, or artist is remembered today. Sparta’s contribution to human civilisation, beyond the military arts, is effectively zero.
More importantly, Sparta’s rigidity made it unable to adapt. Its system depended on a large population of helots (serfs) who outnumbered Spartan citizens by as much as seven to one. When Spartan citizen numbers declined — because the system was so demanding that fewer families could meet the requirements for full citizenship — the entire structure became unsustainable. By the third century BC, Sparta had fewer than 1,000 full citizens and had become a military irrelevance. The society that had controlled everything could not control the one thing that mattered: its own demographic decline.
The parallel with modern attempts to control information is instructive. Societies that prioritise control over openness may achieve short-term stability and even military power. But they sacrifice the adaptability, innovation, and self-correction that come from allowing people to speak freely, challenge orthodoxies, and propose new ideas. In the long run, the open society outperforms the closed one — not because openness is morally superior (though many would argue it is), but because it is functionally superior. Open societies process information better, correct errors faster, and innovate more effectively.
Part 3: The Islamic Golden Age — Openness Ascendant, Then Lost
The Islamic Golden Age (roughly 750–1258 AD) provides one of history’s most dramatic demonstrations of what happens when a civilisation embraces intellectual openness — and what happens when it stops.
Under the Abbasid Caliphate, centred in Baghdad, the Islamic world became the intellectual leader of the planet. The House of Wisdom (Bayt al-Hikma) in Baghdad was the world’s greatest centre of learning. Scholars translated the Greek philosophical and scientific corpus into Arabic, preserving works that would otherwise have been lost. They then built upon this knowledge, making original contributions in mathematics (algebra, algorithms), astronomy, medicine, optics, chemistry, and philosophy that were centuries ahead of contemporary Europe.
This intellectual flowering was not an accident. It was the product of a culture that valued intellectual inquiry and tolerated — indeed, encouraged — debate. The practice of ijtihad (independent reasoning) allowed scholars to interpret religious texts in light of new knowledge and changing circumstances. Theological disputes were conducted openly. Christians, Jews, Zoroastrians, and Muslims studied alongside each other. The Caliph al-Ma’mun (r. 813–833) personally hosted debates between scholars of different faiths and philosophical traditions.
The decline began when this openness was curtailed. The closing of the “gates of ijtihad” — the gradual consensus that independent reasoning on matters of religious law was no longer permissible — is one of the most consequential intellectual events in history. The exact date and nature of this “closing” is debated by scholars, but the effect is not: Islamic intellectual culture shifted from creative inquiry to the preservation and commentary of existing texts. Original research declined. The sciences stagnated. The philosophical tradition withered.
By the time the Mongols sacked Baghdad in 1258, the Islamic world’s intellectual leadership had already been eroding for generations. The Mongol destruction was catastrophic — the House of Wisdom was destroyed, its books thrown into the Tigris — but it was the culmination of a decline, not its cause. The cause was the decision, made incrementally over centuries, to restrict the range of permissible thought.
The Islamic world never recovered its intellectual pre-eminence. The very civilisation that had given Europe algebra, the concept of zero, and the preservation of Aristotle found itself, by the sixteenth century, falling behind a Europe that had embraced the very intellectual traditions that the Islamic world had abandoned: open inquiry, empirical investigation, and the willingness to challenge received wisdom.
Part 4: The Soviet Experiment — Controlling Everything
The Soviet Union represents the most thorough modern attempt to control information. The Communist Party controlled every newspaper, every radio station, every television channel, every publishing house, every school curriculum, and every public lecture. Censorship was not a side effect of the system. It was the system. The entire apparatus of the state was designed to ensure that Soviet citizens received only the information that the Party deemed appropriate.
The results were predictable to anyone who had studied history, and catastrophic for those who had not.
Soviet science, which had produced genuine breakthroughs in the early decades (rocketry, nuclear physics, mathematics), was progressively crippled by ideological interference. Trofim Lysenko’s pseudoscientific theories about genetics — which aligned with Marxist ideology but contradicted actual biology — were imposed as official doctrine from the 1930s to the 1960s. Scientists who objected were imprisoned, exiled, or killed. Soviet agriculture suffered catastrophically as a result, contributing to famines that killed millions.
Soviet economics operated on similarly distorted information. Central planners could not get accurate data about production, demand, or quality because every level of the system had incentives to falsify reports. Factory managers overstated output to meet quotas. Regional officials inflated statistics to please Moscow. Moscow published the inflated statistics as evidence of socialist superiority. The entire economy was running on lies, and the people at the top had no way of knowing it — because they had destroyed the mechanisms (free press, independent auditors, market prices) that would have told them the truth.
The samizdat (self-published) underground demonstrated that even the most comprehensive censorship regime in history could not fully suppress the flow of information. Dissidents hand-typed manuscripts and passed them from person to person, creating an alternative information network that, while tiny, proved that the human desire to communicate freely is essentially impossible to extinguish.
The Soviet Union collapsed in 1991 — not because it was invaded, but because it could no longer function. A system built on controlled information could not process the information it needed to survive. The censorship that was supposed to protect the regime was the very thing that destroyed it, by blinding it to its own failures.
Part 5: The Digital Dilemma
Every previous technology for controlling information — book burning, licensing of printers, state monopoly over broadcasting — was eventually defeated by new technology. The printing press defeated the Church’s manuscript monopoly. Cheap newspapers defeated government licensing. Radio and television defeated state control of the printed word. The internet defeated state control of broadcasting.
AI-powered censorship represents something qualitatively different. For the first time in history, governments have access to technology that can monitor, filter, and suppress speech at a scale and speed that no previous censorship regime could match. China’s Great Firewall, combined with AI-driven content moderation, can identify and remove a social media post within seconds of its publication. India’s three-hour takedown rule, enforced by AI systems, approaches real-time censorship. The EU’s Digital Services Act, backed by algorithmic auditing tools, creates a compliance burden that effectively forces platforms to over-censor rather than risk regulatory penalties.
The question is whether AI-powered censorship will prove to be the exception to the historical rule — whether, for the first time, a censorship technology will be powerful enough to suppress dissent permanently without the self-defeating consequences that destroyed every previous attempt.
Thomas Sowell’s insight about markets applies equally to information: no central authority, no matter how intelligent or well-intentioned, can process information as effectively as a distributed system of free agents making independent decisions. Central planning of information — which is what censorship is — suffers from the same fundamental problem as central planning of the economy: the planners cannot know what they do not know. The information they suppress may be precisely the information they need to make good decisions. The dissent they silence may be the early warning of a problem that will destroy them.
Part 6: What Comes Next
The 2030s will see a divergence between open and closed information societies, with consequences as profound as the divergence between Athens and Sparta.
On one side will be nations that embrace open discourse — including speech that is uncomfortable, offensive, or politically inconvenient. These nations will benefit from faster error correction, more effective innovation, and greater social resilience. They will attract the world’s most talented and ambitious people, who will gravitate toward places where they can think, speak, and create without fear.
On the other side will be nations that use AI to enforce ever-more-comprehensive control over what their citizens can say, read, and think. These nations will achieve a surface stability that looks impressive — orderly, efficient, conflict-free. But beneath the surface, they will accumulate errors, suppress innovation, and lose the adaptive capacity that allows societies to respond to unexpected challenges.
The EU risks becoming a “comfortable Sparta” — a civilisation that is orderly, well-regulated, and intellectually sterile. Its regulatory approach to speech and AI, however well-intentioned, creates a compliance culture that punishes dissent and rewards conformity. The Digital Services Act does not just regulate platforms. It creates an incentive structure in which platforms systematically remove anything that might attract regulatory attention — which means anything controversial, provocative, or genuinely new. The result is a public discourse that is safe, bland, and incapable of producing the kind of radical thinking that drives civilisational progress.
The historical pattern is clear and consistent across millennia: civilisations that control speech decline. They decline not because censorship makes people unhappy (though it does), but because it makes societies stupid. It blinds leaders to reality. It prevents the identification and correction of errors. It drives talented people away. It replaces genuine debate with performance and compliance.
Every civilisation that has chosen the silence of the scribes has eventually been overtaken by one that let its people speak. There is no reason to believe that the twenty-first century will be different.
The only question is which nations will learn from history, and which will repeat it.