Brave New Information World
From censorship to surveillance to self-censorship
In September 2026 the British government announced plans for a new National Centre for Information Defence. Speaking at the United Nations, the Prime Minister described an “industrial-scale assault” on Britain’s information environment by hostile states, Russia in particular, using bots, fake websites and forged newspaper material. The new centre would “detect, attribute and disrupt” such attacks, and Britain needed the “architecture” to respond.
A democracy has legitimate reasons to protect itself against covert foreign interference. But the word architecture matters. In June the government’s media green paper, Watch This Space, floated legislation requiring platforms to make news from designated “trustworthy” providers prominent and easily discoverable, arguing that such content supports common references and a “shared cultural identity”. In Europe, the Democracy Shield sets out to safeguard the integrity of the information space. None of these measures amounts to the state declaring what citizens may believe. Together, though, they show that the information environment itself has become an object of governance.
The message around the message
What struck me was how I encountered the announcement. Not through the speech or a policy paper, but through my LinkedIn feed, which quickly filled with endorsements from people working in security, technology and policy. One expert argued the new centre should be “always on, always learning” and able to operate across “all the information spaces that matter”.
Which spaces are those, and who decides they matter? Where does a hostile operation end and ordinary disagreement begin?
My feed is no referendum on public opinion. It is algorithmically selected from my network and interests. But that is precisely the point. I was not encountering public opinion. I was encountering an information architecture, and inside it the proposal arrived already surrounded by expertise, endorsement and social proof. No coordination is needed for that to happen. Government, think tanks and professional networks can converge around a framing without anyone directing the whole. The architecture delivers not only a proposition but a sense of what every sensible person already thinks about it.
That matters, because human beings learn environments.
Censorship, surveillance, self-censorship
In my System & Self work, the relationship runs both ways. The system looks at the self, and the self looks back, learning what the system notices, rewards and treats with suspicion. In time the self anticipates the system, and eventually carries it inside.
I increasingly see the past decade as three overlapping movements. Around 2016 the argument was about censorship: moderation, deplatforming and who set the boundaries of the digital public square. By the pandemic years, surveillance had become the visible concern, with tracking, health data and digital identity making individuals ever more legible to institutions. Now, I think, the defining problem is becoming self-censorship, and it may carry us towards 2030.
Censorship says: you cannot say this.
Surveillance says: we can know that you said this.
Self-censorship says: perhaps I had better not say this.
Research supports the worry. Elizabeth Stoycheff found that perceptions of government surveillance can chill the expression of minority political views. A 2026 University of Zurich experiment that reminded 827 people about dataveillance over six weeks found it reduced their willingness to do entirely lawful things online, from expressing opinions to seeking and disclosing information. Perhaps I won’t search for that. Perhaps I won’t ask that question.
No instruction is required. The self has learned the environment well enough to intervene on itself.
Orwell outside, Huxley inside
We usually reach for Orwell here, and understandably. The system in Nineteen Eighty-Four is spectacularly visible. It watches, commands and punishes, overwhelming the self from outside. Huxley’s Brave New World goes further in. And I have always said that the system conditions and rewards, producing people so well adapted to their environment that little overt censorship is ever needed.
Britain in 2026 is not Brave New World. The useful point is that power need not work against people. It can work through the environments in which judgments and identities are formed, and those environments usually present themselves as helpful: safer, more trustworthy, more resilient. Benevolent intention and architectural power sit comfortably together. The green paper’s own case rests on information architecture shaping society; that is why it purports to be worth doing.
So the question is no longer only who controls the information. It is what happens to the self that learns to live inside the system organising it.
Somewhere in that movement, censorship completes its transformation. The censor no longer needs to stand outside us.
The censor has become the self.