The “censorship industrial complex” (CIC) is what its backers describe as “a network of ideologically-aligned governmental, NGO, and academic institutions” that suppresses populist and right-wing online speech. It was popularised by commentators on the American right and is now part of US government policy. Supporters point to cases such as officials in President Joe Biden's White House pressing Facebook to remove posts about Covid vaccines in 2021. Critics, including the researchers Renée DiResta and Dean Jackson, call it a propaganda frame. In their view, it takes routine, voluntary work and presents it as secret censorship. During the 2020 US election, for example, researchers at Stanford University sent social media companies examples of posts they believed were spreading false information about the election.

The argument has now reached Britain, where critics including Reform UK say the Online Safety Act and the new National Centre for Information Defence threaten free speech. The Trump administration has also used the term against Europe, using it to accuse European campaigners and officials of pressuring American tech companies into censoring posts. Yet the evidence suggests that what most people see online is shaped less by governments or researchers than by the recommendation algorithms platforms use to keep people scrolling.

Where the phrase “censorship industrial complex” comes from

The phrase borrows from President Dwight Eisenhower's 1961 warning about the “military-industrial complex”. It entered wide use in March 2023, when the writers Michael Shellenberger and Matt Taibbi testified before a US House of Representatives committee about the “Twitter Files”, a cache of internal Twitter documents released after Elon Musk bought the company. Shellenberger titled his written testimony The Censorship Industrial Complex, and Taibbi, in his own testimony, credited him with the phrase.

However, neither man was the first to use it. Others had used it earlier, and the online commentator Mike Benz did more than most to spread it, according to an investigation by MIT Technology Review. The investigation follows his claims from videos and posts into congressional hearings, and from there into government policy.

The case for: what proponents of CIC point to

Much of the case rests on the US government leaning on social media companies over what their users could post. In 2021, officials in President Joe Biden's White House repeatedly pressed Facebook to remove or limit posts about Covid vaccines. A 2024 report by Republicans on the House Judiciary Committee said this campaign “targeted true information, satire, and other content”. Twitter's decision in 2020 to stop users sharing a New York Post story about Hunter Biden's laptop is often cited too, and its co-founder Jack Dorsey later called it a mistake.

The second strand of the argument involves universities. During the 2020 election, researchers at Stanford's Election Integrity Partnership reported posts they considered misleading to the platforms, which could then label or remove them. The Twitter Files journalists, in a brief to the Supreme Court, argued that government agencies had “outsourced” censorship to projects like this one, getting round the First Amendment, which stops the US government from censoring speech itself.

The case against: what critics of CIC say

Critics accept that some officials and politicians have contacted social media platforms, but dispute claims that that this has added up to a coordinated system of censorship. The question reached the US Supreme Court in 2024, after two states and five individuals sued the Biden administration in Murthy v. Missouri. The court threw the case out by six votes to three, finding that the challengers could not show the government's contacts had censored them. That left open whether officials had crossed a line, but the judges noted that “the platforms had independent incentives to moderate content and often exercised their own judgment”.

A 2026 investigation by MIT Technology Review into whether the alleged network exists found that platforms did remove and limit posts, mostly under their own rules, but turned up no proof of coordination. Renée DiResta, who worked at the Stanford Internet Observatory, one of the main targets of the claims, goes further. In a peer-reviewed paper with Dean Jackson, she argues that the CIC label was used to justify cutting the US government's work against election disinformation. The Observatory itself wound down its election research in 2024 after various lawsuits and congressional subpoenas that, according to Platformer, forced Stanford “to spend millions of dollars to defend its staff and students”.

How the idea became US policy

Within two years, the phrase had travelled from congressional testimony and online commentary into the language of the US government. On his first day back in office in January 2025, President Donald Trump signed an executive order declaring that “government censorship of speech is intolerable in a free society”. Meta had already shifted its approach to content moderation. A fortnight earlier, its chief executive, Mark Zuckerberg, had ended fact-checking on Facebook and Instagram in the US, arguing there had been “too much censorship” while conceding that the platforms would now “catch less bad stuff”.

In March 2025 the Senate held a hearing under the title “The Censorship Industrial Complex”. Senator Peter Welch, the panel's senior Democrat, rejected its premise outright, saying “My view is that facts don't support that allegation.” A month later the Secretary of State, Marco Rubio, shut the State Department office responsible for exposing disinformation campaigns by foreign states (our guide to FIMI explains how such campaigns work). Rubio set out his reasons in an article headlined “To Protect Free Speech, The Censorship-Industrial Complex Must Be Dismantled”.

The administration has since carried the argument abroad. In December 2025 the State Department barred five Europeans from entering the US, branding them “agents of the global censorship-industrial complex”. One was Clare Melford, who runs the Global Disinformation Index, a London-based organisation that advises advertisers on which websites spread disinformation. In March 2026 the Justice Department settled the original Missouri lawsuit. Under the agreement, approved by a federal judge, the Surgeon General, the Centers for Disease Control and Prevention, and the Cybersecurity and Infrastructure Security Agency may no longer threaten social media companies with punishment to force the removal of posts protected by the First Amendment. Because the administration agreed with the plaintiffs and settled, no court has ruled on whether the Biden-era pressure broke the law.

Is censorship illegal in the UK?

Britain has no equivalent of the US First Amendment. Free speech is protected by Article 10 of the European Convention on Human Rights, part of UK law through the Human Rights Act 1998, but it is not absolute. The law can restrict it, for example to stop threats or incitement to violence.

At the centre of the British dispute is the Online Safety Act 2023, which obliges social media companies to take down illegal content, from terrorist propaganda to threats of violence, and to shield children from material such as pornography, on pain of fines from the regulator, Ofcom. What the Act does not explicitly set out to do is police the truth. The government has confirmed that false or misleading posts were deliberately left outside its scope, in line with a cross-party committee of MPs, which concluded in 2025 that “neither government nor private companies should be arbiters of truth”. Ofcom's investigation into X reflects that narrower focus, and according to Reuters centres on illegal hate speech, terrorist material and sexual deepfakes generated by Grok, the platform's AI chatbot.

Critics say the Act still does too little to protect free speech, a view supported by a 2024 legal study. US officials told Ofcom in 2025 that they were “concerned about freedom of expression in the United Kingdom”, according to the Guardian. In September that year Nigel Farage, the leader of Reform UK, raised the same worry at a US Congress hearing called “Europe's Threat to American Speech and Innovation”.

The argument intensified in September 2026, when the Prime Minister, Andy Burnham, announced a National Centre for Information Defence to counter disinformation from hostile states. Farage said it “will be used to clamp down on free speech”, although he accepted that “there are bots online” run from Russia and elsewhere. The government has not yet said what powers the centre will have or who will oversee it. Sophie Williams-Dunning of the defence think tank RUSI told The Record that there is “a real difference between a centre inside the Cabinet Office and an agency accountable to Parliament.”

Who actually controls what you see online?

The censorship debate fixes on what gets taken down, yet for most people what they see online depends far more on what platforms choose to promote. Ofcom's 2026 news survey found that 73% of UK adults get news through social media, search engines or other online services, where what appears is “ often driven by recommendation algorithms”. The cross-party committee of MPs described these algorithms as “integral to the advertisement- and engagement-driven business models” of the platforms, which make money by keeping people scrolling. Even removals are largely the platforms' own decisions. An analysis of a day's worth of EU moderation data by the US think tank ITIF found that 97% of the content was detected by automated systems, and that platforms removed most posts on their own initiative, for breaking their own rules, rather than waiting for takedown notices or government orders.

Those rules can be rewritten overnight. After Elon Musk bought Twitter, now X, the company dropped its rules on Covid and election misinformation, according to the Alexander von Humboldt Institute for Internet and Society. Algorithms can also tilt what users see. A study for the Finnish innovation fund Sitra found that 58% of the political posts shown to 24 test accounts, set up as young people in Finland, France and Romania, were right-wing. Whether that changes minds is less clear, and a large 2023 experiment on Facebook found that altering people's feeds had little effect on their political views. Neither side of the censorship debate can fully check what platforms are doing, because their transparency reports, in the words of the Anti-Defamation League, “continue to be anything but transparent”.

What this means for journalism and trust

Fact-checkers and disinformation researchers, whose job is to check what is true, have themselves been cast as part of the “censorship industrial complex”. Meta dropped its US fact-checkers, Stanford's researchers faced lawsuits, and the head of a London-based disinformation research group was barred from the US. When people in power describe this work as censorship, it appears to become harder for the public to know whom to trust.

The Reuters Institute found that in 2026 only 30% of people in the UK trust most news most of the time, five points down on the previous year and below a global average of 37%. 50% avoid the news at least sometimes, and 77% worry about what is real or fake online. Journalism is far from dead, though. The BBC still reaches 77% of UK adults, and Ofcom found that 65% rate broadcast TV news as trustworthy, against 39% for news on social media.

Frequently asked questions

What is the “censorship industrial complex”?

It is a term for an alleged network of governments, researchers, charities and tech platforms said to work together to suppress lawful speech online. Supporters of this theory point to US officials pressing platforms over Covid posts. Critics say there is no proof of a coordinated network. The US Supreme Court did not settle the question when it heard a case on it in 2024.

Who coined the term “censorship industrial complex”?

Michael Shellenberger and Matt Taibbi popularised it at a US congressional hearing in March 2023, and Taibbi credited the phrase to Shellenberger. Others had used it earlier, according to MIT Technology Review.

What does “industrial complex” mean?

The phrase comes from President Eisenhower's 1961 farewell address, which warned against “the acquisition of unwarranted influence” by the “military-industrial complex”, the alliance of the armed forces and the arms industry. The censorship version applies the same idea to governments, researchers and tech firms.

Is the censorship industrial complex real?

The contacts behind the claim were real. US officials did press platforms, including Facebook, over Covid posts, and in 2026 the Justice Department agreed to bar several agencies from threatening social media companies. What has not been shown is a coordinated network deciding what people can say. The US Supreme Court noted that platforms “often exercised their own judgment”, and a 2026 investigation by MIT Technology Review found no proof of coordination. Most removals are made by the platforms under their own rules.

Is the National Centre for Information Defence a “Ministry of Truth”?

Critics use that phrase, but the centre's powers have not yet been set out. France's equivalent agency, Viginum, avoids judging whether individual claims are true and focuses on foreign operations, partly to avoid such comparisons, according to The Record.