There’s a phrase for what’s happening: “misinformation about misinformation.” It was coined by five academics led by Ceren Budak of the University of Michigan, in a paper in Nature in June 2024, to describe the gap between what public figures claim about online falsehood and what the evidence actually shows. That phrase could serve as the subtitle for what the Labour Government is now proposing in Britain. The Culture Secretary, Lisa Nandy, has published a Green Paper called Watch this Space that would force social media and video-sharing platforms, by law, to give priority to “regulated public service media” like the BBC, so that trusted outlets are “seen and heard in the fierce battle against mis and disinformation.” Lucy Powell, Labour’s Deputy Leader, wants election law changed so that online platforms are held to the same accuracy and balance standards as broadcasters—a plan that would require a huge new surveillance and censorship machinery that does not yet exist. Meanwhile, after amendments to the Online Safety Act, the regulator Ofcom has told platforms to create protocols for restricting content during a “crisis,” defined vaguely as an extraordinary situation with a serious threat to public safety. Taken together, these ideas represent the most far-reaching machinery for controlling public speech that Britain has seen in peacetime since the seventeenth century.
The entire architecture rests on a single document: a 136-page “evidence review” on mis- and disinformation, commissioned by the Government Office for Science from a team at Sheffield University and published in May. It was supposed to provide the “full evidence base” that ministers admitted they had previously lacked. Its tone was apocalyptic. It declared that online fake news is “widely recognised as a systemic national risk,” that algorithms and business models amplify falsehood, and that “system-level interventions are needed.” It recommended statutory “quality signals” in ranking content, a state-linked “database of false and misleading narratives,” and official “prebunking” in schools and public communications. But when you look closely at how the review was put together, it falls apart. The authors did not actually read the studies they cite. They relied on summaries generated by large language AI models of 228 academic papers they had never opened. They also used monthly bulletins from the European Digital Media Observatory (EDMO), a fact-checker coordinating body. Those bulletins are not measurements of misinformation. They are questionnaires asking fact-checkers what subjects they happened to work on that month. So when the review says health misinformation was “detected” every month, it only means fact-checkers chose to check that topic. It says nothing about how many misleading posts existed, how many people saw them, by what standard they were judged false, or whether anyone believed them. The review’s authors and its eleven “expert” reviewers were not equipped to make such judgments anyway. There was no climate scientist, no economist, no clinician, no immunologist, no epidemiologist, and no migration scholar—yet the report confidently pronounces on misinformation about climate, EU financial policy, transgender sportspeople, and migrant entitlements. Instead, it included long-serving members of what has been called the censorship industrial complex: Clare Melford of the Global Disinformation Index, Chris Morris of Full Fact, Paula Gori of EDMO, and Bob Ward of the London School of Economics, whose career largely consists of complaining about journalists who question climate and energy policy. These are partisans in the very contest the review was supposed to adjudicate, and their own work is cited repeatedly in its text. They were literally marking their own homework.
Melford’s record shows what that means in practice. The Global Disinformation Index (GDI), which she co-founded in 2018, does not correct falsehoods. It tries to defund publishers by pressuring advertisers. It rates news websites for “disinformation risk,” compiles a “dynamic exclusion list,” and supplies that list to ad agencies and clients, so that outlets it dislikes are starved of revenue. In 2020, at the beginning of the pandemic, the GDI attacked Senator Tom Cotton for suggesting on Fox News that Covid-19 might have escaped from a laboratory in Wuhan. According to the GDI, his words gave a “debunked conspiracy theory” authority and validation, and advertisers had to withdraw from sites that echoed it. That position has not aged well. Many experts, governments, and intelligence agencies now believe the lab leak was not a conspiracy theory at all, and that the virus very likely did escape from the poorly shielded Wuhan lab. The GDI did not stop at origins. In 2021, it denounced a Spanish-language site that reported that a third of British deaths from the Delta variant were among the vaccinated. It called this “clearly untrue” and persuaded the restaurant chain Chipotle to pull its advertisement from beside the article. But Public Health England’s own technical briefing, published on 18 June 2021, recorded 73 deaths in England among people confirmed to have the Delta variant. Of those, 34 were unvaccinated, ten had received one dose, and 26—36 percent—had been fully vaccinated for more than a fortnight. More than half of the dead had received at least one dose. So the site was right, and the GDI was wrong. Meanwhile, the GDI and its allies have been quietly redefining disinformation to include not just falsehoods but “adversarial narratives”: content that may be perfectly accurate but that, in their judgment, sets people against an individual, an institution, or “the science.” In 2024, the GDI told UnHerd that the site would stay on its exclusion list because it published gender-critical writers like Kathleen Stock, whose views are protected under British law. For UnHerd, the consequences were severe: its advertising revenue fell to between 2 and 6 percent of what its audience would normally command. Another component of the censorship lobby, the Centre for Countering Digital Hate—co-founded by Morgan McSweeney, who now works in Downing Street—campaigns against the “new climate denialism,” a category that includes not denying physics but simply questioning Britain’s Net Zero energy policy, or expressing doubts about heat pumps and electric vehicles. And it is a matter of public record that between 2019 and 2023, the Foreign Office paid the GDI grants worth £2.6 million. Given all this, it is hardly surprising that the GO-Science review reached the conclusions it did. Its proposed state-linked “database of false and misleading narratives,” refreshed monthly, is essentially the GDI’s bank of disagreeable narratives with a government seal of approval. Having commissioned an evidence base, ministers were given a predetermined manifesto.
The review’s claim about algorithms is even weaker. It asserts that online platform algorithms are exclusively profit- and engagement-driven, and therefore foster hatred and division. That sweeping conclusion rests on a single citation: a 2025 paper from staff at Universitas Muhammadiyah Buton, a small private institution in Southeast Sulawesi, Indonesia, ranked by UniRank as the 433rd best of 562 universities in Indonesia, and the 10,370th best in the world. That is not a typo. The most significant research in the world’s leading journals contradicts the review at every turn. The Nature paper by Budak and her colleagues documents “a pattern of low exposure to false and inflammatory content that is concentrated among a narrow fringe with strong motivations to seek out such information.” Visits to 490 websites deemed “untrustworthy” by fact-checkers made up just 5.9 percent of Americans’ clicks on news sites. When television is added to the mix, “fake news” amounts to roughly 0.15 percent of the average media diet. The GO-Science review appeared two years later and did not cite this paper at all. On algorithms specifically, the largest study was conducted by researchers at Twitter itself and published in PNAS in 2022, covering seven countries and millions of accounts. It found that the algorithm did amplify right-of-centre legislators somewhat more than left-of-centre ones, but when it came to extremism, “contrary to prevailing public belief, we did not find evidence to support this hypothesis.” Far-right and far-left parties in France, Spain, and Germany were amplified less than moderate parties, not more. A landmark experiment published in Science in 2023 switched tens of thousands of Facebook and Instagram users from an algorithmic feed to a simple chronological one for three months. Their political attitudes did not measurably change. And the algorithmic feed actually served them less content from untrustworthy sources, not more. Studies of YouTube found that consumers of extremist videos were overwhelmingly subscribers who went looking for them, not innocent viewers led astray by recommendations. Budak and her colleagues drew the obvious conclusion: although algorithms shape what people see, the evidence suggests that, on average, they tend to push users toward more moderate content, and offer extreme content mainly to those who have already sought it out. None of this means the internet is harmless. It means the panic is not proportionate to the problem, and the panic is driving the policy.
The same pattern holds for the claim that misinformation causes serious harm, such as vaccine hesitancy. The most substantial study of that question was published in Science in 2024. It combined survey experiments with Facebook exposure data covering 233 million American users. It found that content flagged as false by fact-checkers had almost no aggregate effect on vaccination intentions, for a very simple reason: hardly anyone saw it. Such material attracted just 0.3 percent of the 2.7 billion views of vaccine stories in the dataset. What did depress vaccination intentions was unflagged “vaccine-sceptical” content that contained no falsehoods at all. That accurate but unsettling material had 46 times the aggregate impact of everything the fact-checkers had flagged. The single most influential item was not a conspiracy theory. It was an accurate report in the Chicago Tribune headlined “A healthy doctor died two weeks after getting a COVID vaccine; CDC is investigating why.” It was seen by more than 50 million people—over six times the views of all flagged vaccine misinformation put together. Advocates of regulation cite this study as proof that moderation must go further. But it proves the opposite. The material with the measurable effect was truthful reporting by an established metropolitan newspaper—exactly the kind of “trustworthy” journalism Nandy’s Green Paper proposes to amplify. A regime aimed at misinformation would have missed it completely. A regime capable of suppressing it would have crossed a dangerous line: from correcting falsehood into suppressing accurate news that officials found inconvenient. That is the choice the censorship lobby rarely makes explicit, though Melford came close in 2021 when she explained that her organisation had moved beyond checking facts because, in her words, “something can be factually accurate but still extremely harmful.” If misinformation really were the threat ministers describe, you would expect its footprint to show up in what people believe. But a study published in PLOS ONE in 2022 examined national US surveys of 46 separate conspiracy beliefs, repeated with identical wording since 1966, plus polling on six theories in six European countries, including Britain. In no instance did the researchers find systematic evidence of an increase in conspiracism. Of the 46 American beliefs, significant declines outnumbered increases by 15 to six. In Europe, only one theory had gained ground in a single country. Interestingly, the belief that had grown most was the idea that conspiracy theories themselves are out of control: 73 percent of Americans thought so, a conviction the authors attributed to alarmist statements by politicians and the press. Online conspiracy content, they concluded, might not persuade people so much as reinforce views they already held.
Britain has been here before. In July 1637, the Star Chamber of Charles I decreed that nothing could be published without a licence from approved authorities, with enforcement delegated to a private body, the Stationers’ Company. The forbidden categories—“seditious,” “scandalous,” “offensive”—meant whatever the licensers wanted them to mean. The proposals of 2026 are not an exact echo, but they are close enough to be unsettling. Exalted legal status for approved publishers, emergency powers over content, and official programmes to teach children and adults which voices to trust are the closest peacetime relatives to that regime in nearly four centuries. The Star Chamber’s measures failed, and Charles was beheaded twelve years later. Seven years after the King’s decree, John Milton inaugurated England’s free speech tradition with Areopagitica. His answer to the licensers still stands. If we believe in truth, Milton wrote, “we do injuriously, by licensing and prohibiting, to misdoubt her strength. Let her and Falsehood grapple; who ever knew Truth put to the worse, in a free and open encounter?” The Government’s case rests on a review that read nothing, measured nothing, and ignored the research that contradicted it. It relies on people who have built careers out of treating accurate reporting as dangerous, and who have been paid by the state to do so. The most far-reaching apparatus to restrict peacetime speech since the ousted Stuart king deserves far better than that. On the evidence, it amounts to exactly what the academics said: misinformation about misinformation.