UnHerd, the Britain-based publication I lead, published an investigation on April 17 into a transatlantic organization called the Global Disinformation Index. We revealed that, having received money from the U.S. State Department, as well as the British, German and European Union governments, the GDI issues what amount to blacklists of news publications, on highly tendentious grounds, that online advertising exchanges then consult and can use to justify turning off ad revenue.
With worries about the rise of “disinformation” in recent years, various projects were launched in the United States, Britain and elsewhere, many no doubt with good intentions, to combat disinformation’s deleterious effects on democratic values. What has emerged, though, is an opaque network of private and government-supported enterprises that appear intent on censoring political views they find unpalatable.
Just last month saw the U.S. launch of a new effort billing itself as combating disinformation — but its political agenda was unmistakable. The American Sunlight Project is the brainchild of Nina Jankowicz. She headed President Biden’s Disinformation Governance Board for the three weeks of its existence in 2022, until it was abandoned under a barrage of criticism for its Orwellian name and unclear mission.
The American Sunlight Project’s goal, as the New York Times reported, is to fight back against “what she and others have described as a coordinated campaign by conservatives and others to undermine researchers, like her, who study the sources of disinformation”. In other words, the newest addition to this expanding bureaucracy is an anti-anti-disinformation unit — built to defend the fact-checking fraternity against attacks. Jankowicz has become a pugnacious presence on social media, seemingly offering herself as a public spokesperson for what increasingly looks like a political project.
Determining the extent of the damage done to media properties in recent years by self-appointed disinformation monitors is difficult because their influence on the complex machinery that serves online advertising is hard to measure. It is even unclear which groups’ evaluations are heeded in this murky system.
But I can attest that UnHerd has been substantially affected: Though NewsGuard, another disinformation ratings organization, gives us a trust score of 92.5 percent (five points ahead of the New York Times), the GDI at some point last year mysteriously placed us on their “dynamic exclusion list” of publications that supposedly promote disinformation and should be boycotted by advertisers. As a result, tech giant Oracle, which has a relationship with the GDI, provided a poor “brand safety” rating to our ad agency, and we received only a tiny fraction of the ad revenue the agency had predicted for our audience. Thankfully, we are primarily subscriber-funded, but for smaller publications more reliant on ad revenue, this would be a death knell to their business.
What did UnHerd do to provoke the GDI’s disapproval? After repeatedly asking the organization for an explanation, we eventually got an answer: “Our team re-reviewed the domain, the rating will not change as it continues to have anti-LGBTQI+ narratives. … The site authors have been called out for being anti-trans. Kathleen Stock is acknowledged as a ‘prominent gender-critical’ feminist”.
They did not point to any factual errors — their complaint was with the viewpoints of some of our contributors. In addition to decrying Stock, a prominent British philosopher and co-director of the Lesbian Project, the GDI email pointed to Julie Bindel, a lifelong campaigner to stop violence against women, and Debbie Hayton, who is transgender. Apparently the GDI equates “gender-critical” beliefs, or maintaining that biological sex differences exist, with “disinformation” — despite the fact that those beliefs are specifically protected in British law.
The GDI similarly targets other issues — such as climate change and the origins of the coronavirus — that are more properly the subject of robust debate, not matters of “disinformation” if a writer simply has a viewpoint that GDI disapproves.
When the index was originally set up, in 2018, it defined disinformation as “deliberately false content, designed to deceive”. On this basis, you could see the argument for having fact-checkers to identify the most egregious offenders and call them out. But mission creep has set in at the GDI. It has since come up with a definition of disinformation that encompasses anything that deploys an “adversarial narrative” — stories that might be factually true but pit people against one another by creating “a risk of harm to at-risk individuals, groups or institutions” — with institutions defined as including “the current scientific or medical consensus”.
GDI co-founder Clare Melford explained in a 2021 interview at the London School of Economics how this expanded definition was more “useful”, as it allowed the GDI to go beyond fact-checking to flagging any online material the organization deemed “harmful” or “divisive”.
In December 2022, the GDI issued a report listing the 10 U.S. publications that posed the most “risk” of promoting “disinformation”. It looked distinctly like a list of the country’s most-read conservative websites, including the New York Post and RealClearPolitics.
In December last year, two publications on the GDI list, the Daily Wire and the Federalist, teamed up with the attorney general of Texas to sue the State Department for helping fund GDI and NewsGuard. In recent years, GDI has received hundreds of thousands of dollars in backing from the State Department and other government-related entities. The British government is an even heavier backer: From 2019 to 2023, the Conservative government — perhaps to the astonishment of Tory voters, if they had been aware — directed about $3.2 million to the GDI, which also is backed by George Soros’s Open Society Foundations and other liberal organizations.
The de facto alliance between government and groups working to defund disfavored publications — a sort of state censorship laundering arrangement — is particularly alarming. Congress is awakening to the problem: It sent a message on this front with the 2024 National Defense Authorization Act, barring the Defense Department from placing military-recruitment advertising in publications utilizing GDI, NewsGuard or “any similar entity”.
But as we have seen at UnHerd, the unaddressed problem with these disinformation referees is how their rulings affect online ad services themselves, not just advertisers, with the power to throttle revenue to publications simply for ideological reasons.
It isn’t even clear that the bosses of big tech companies understand the extent that their own organizations have become entangled in this movement. A spokesman for Oracle last year announced that they would be ending their relationship with the GDI on free speech grounds, but our research shows that Oracle is still collaborating via their ad tech platform, Grapeshot. One wonders: Is Oracle’s founder and chairman, Larry Ellison, a Republican donor, aware of this? Meanwhile, Elon Musk responded to our investigation by saying on X that the GDI should be “shut down, with recriminations for the miscreants”, apparently unaware that his own company, X, is collaborating with the GDI via X’s partnership with Integral Ad Science for brand safety information.
There is no doubt that an open, free internet means bad information can travel like never before. But attempts to impose censorship of political speech under the apparently innocuous banner of combating “disinformation” — whether the projects are highly publicized, like the American Sunlight Project, or secretive and pseudo-technical, like the Global Disinformation Index — amount to a much greater risk to a functioning democracy. Not only does censorship not work, but it also adds fuels to the flames of division and paranoia. Next time you hear someone casually use the word “disinformation”, be skeptical: They might well be making the problem worse.
Freddie Sayers is the editor in chief of UnHerd.