How the U.S. government built the anti-disinformation field

Launching a database of the U.S. federal government's anti-dis/misinformation funding

Today, liber-net is launching a searchable database of almost 900 U.S. federal government awards to counter mis-, dis-, and malinformation (MDM) and other content moderation initiatives, covering the period from 2010 to 2025.

We gave the Free Press a preview, and today Gabe Kaminsky and Madeleine Rowley published a story based on the database, focusing on the more than 600 awards made under the Biden administration, though our data shows the “anti-disinformation” industry really took off under Trump 1.0 and then accelerated radically under Biden.

The Free Press spoke with more than dozen government agencies for that story and noted that “since then, federal officials have terminated at least several dozen programs related to misinformation and disinformation, according to documents and interviews.”

Since October of 2024, liber-net has been researching federal MDM funding to inform a set of policy proposals for how the U.S. government-funded wing of the Censorship-Industrial Complex might be dismantled. Ours was a rod-and-spear-style approach to fishing out the censors to identify who exactly was funding what. That method was quickly superseded by DOGE-style dynamite fishing. February in particular saw a massive USAID pile-on that over-inflated the agency’s role in the Complex, as important as USAID was.

After reviewing almost 1100 awards, we included almost 900 in the database. The total value of those awards is roughly 1.5 billion U.S. dollars.

All this sits against the back-drop of the deteriorating Trump 2.0 record on free speech. While there seems to be less government involvement in online speech, the snatching of students off the street (often digitally assisted) or micro-management of private university policy should make anyone wary - such approaches could also easily cross into the online realm. Breadth over accuracy seems to be the rule of the day and it is tarnishing the genuine need for reforms.

Rather than using AI or other tech systems, we personally scoured USASpending (the main database for tracking historical grant, loan, and contract data), the Federal Audit Clearinghouse (a repository of standardized Single Audits from organizations receiving federal funds), grants.gov (current opportunities), sam.gov (registration records, which include contract award data), and a variety of agency-specific databases such as nsf.gov/awardsearch (National Science Foundation-specific), reporter.nih.gov (National Institutes of Health), defense.gov/News/Contracts (Department of Defense), and foreignassistance.gov (State Department/USAID-specific).

Whilst some amount of coding and machine learning can be helpful, there is only so much useful information that a dehumanized approach can produce. Knowing the players, the language, country contexts and so much more is key to understanding what awards are actually problematic, or at least controversial.

To be clear, this isn’t a database of censorship initiatives, though many did seek to censor. The headline isn’t “900 government grants doled out to censor Americans and the world,” but is instead, “How the U.S. federal government became the lead player in developing the anti-disinformation field.” (As per the subtitle) Many of these projects may be legitimate attempts to counter real synthetic influence efforts. The problem is that a large cohort cynically used such claims to sweep up and squash legitimate dissent, calling the field into question more broadly.

Nefariousness or otherwise, the projects share a general world-view of an untrustworthy and ignorant public bent on public disorder held back only by self-appointed sages and counter-disinformation warriors.

We’ve given each of the awards between one and five red flags. Five red flags are for projects that sought to actively remove (or build large-scale systems for removing) content from the internet, or that involved a high level of surveillance on citizens’ speech. For example, we gave five red flags to a University of Illinois project that sought to “track locations, people, and organizational affiliations of dubious COVID-19 information.” “Dubious” meaning anything that contradicted what the government (in this case the Centers for Disease Control and Prevention) said.

In contrast, we gave a one-flag rating to a National Science Foundation (NSF) award to the University of Mississippi’s School of Journalism and New Media which sought to leverage machine learning to detect deep fakes. This project didn’t appear to seek to tilt the political scales. It did, however, use language that indicated it was embedded in the broader “anti-disinformation” ideology and was part of a broader NSF program that included a host of dubious projects.

What we excluded were grants that seemed innocuous, not linked to recent topical controversies, or that used language signalling that they were not part of the dominant “anti-disinformation” ideology that prioritised top-down moderation of content that strayed from officialdom. A National Institutes of Health (NIH) grant to study E-cigarette-related Nicotine Misinformation on Social Media is an example of the kind of award we didn’t include.

You can read more about our methodology here.

What can you do with the database? You can get as specific as looking for grants of over $1m to universities in Kazakhstan funded by the State Department for Covid fact-checking in 2021 - if that’s your thing. Our aim is that it serves as a resource for media organizations, like the Free Press, to dig out stories and bring further transparency and accountability to the “anti-disinformation” field. Please have at it.

The database isn’t complete. There are definitely awards out there that we haven’t identified (if you know of any please contact us) either because we missed them, or in many cases because the government has not publicly published them. In addition, many awards are actively being paused or cancelled. We are going to try and keep it as up-to-date as possible. That said, we're confident the sample size is large enough to show some very clear trends.

As the activities in a grant were not always clear we have erred on the side of caution and only put in those we could be fully confident did involve some level of MDM or content controls activity. That said, some of those grants may well have engaged in such activities, we just can’t be sure. In addition, often only a portion of the grant is for MDM-type activities, however we don’t know which portion and as such have to count the full value of the award.

Not included are actual disinformation campaigns run by the US Government, such as the $493 million given to General Dynamics IT to conduct an actual anti-vax disinformation campaign against the Chinese Sinovac vaccine during Covid, though many of the grants in the database could be considered to be spreading disinformation, rather than countering it, their stated intention at least is to counter it.

Below is an initial series of graphs covering 2016-2024 that show where the money was coming from.

The first is year-on-year funding that includes a massive $979 million Department of Defense (DoD) grant to military contractor Peraton (also reported in the Twitter Files) in 2021 to provide services to U.S. Central Command to “counter misinformation.”

Here is what the year-on-year funding looks like without the Peraton award:

A clear trend is visible which maps onto a commonly-stated assumption: that this work grew rapidly after the Trump 1.0 election, and went gang-busters under Biden and the Covid era, particularly in 2021.

Another way to look at the funding is by the number of individual grants and contracts awarded each year:

Again, a clear trend can be seen. Funding starts to take off in the wake of Trump 1.0 and Brexit, accelerates dramatically in 2020 (likely Covid related), and then increases again in 2021 when Biden takes office.

This roughly correlates with other analyses such as this one tracking the number of mis/disinformation academic studies.

Who was handing out the money? These are the top funders, firstly with that massive DoD Peraton award skewing things:

Take Peraton out and you get:

USAID is clearly the second largest funder by pure dollar amount, but it is still blown away by DoD spending, and the other government agencies combined total more than USAID.

Qualitatively, if you look at awards we have rated with five red flags, the worst offender is the NSF, particularly a program called Convergence Accelerator: Track F.

Looking at the number of awards by U.S. government department also complicates the picture:

You’ll see that the State Department was far and away the most prolific, making 629 awards (more than two-thirds of the entire database), in some cases issuing grants valued at a mere $727. That $727 went to the Atlantic Council for a counter-disinformation workshop in Europe in 2022. The Atlantic Council had revenue of nearly $55m in 2022 - did they really need another $727 from taxpayers?

These 600+ awards represent a huge amount of State Department time, money, and effort that belies the total $57 million we could find. In contrast, we could find only 30 relevant USAID awards in the same period, though they were of a much higher value. Much of the State Department funding was seemingly part of the Building Civic Resilience to the Global Digital Information Manipulation Challenge program. And of course, the infamous Global Engagement Center also sat within the State Department. Their 2024 budget alone (still largely a black box) was $61 million, a significant part of which went to “counter-disinformation” efforts. However, the funding appears to reach far beyond these two programs.

Overall, the key players appear to be the Department of Defense, the State Department, USAID, the National Science Foundation, the National Institutes of Health, and the Centers for Disease Control and Prevention. While there are significant private funders like Craig Newmark, the Knight Foundation, and Omidyar Networks, as well as important EU funding, the U.S. government funding likely overshadows them and has arguably been the biggest driver in developing the anti-disinformation sector.

In the future, we plan to look into EU, private foundation, and other government money to see just how they all stack up. This is particularly important as “anti-disinformation” leaders seek to rebuild these programs from an EU/UK base.

Next week we’ll release a number of other visualisations that focus on where the money went - regions, countries, activities, topics, and more.

Lastly, if you are an academic and are interested in working with this data to produce a research paper, please reach out. In a sane world, there would be journals out there that would love a critical meta-analysis of the sector. The anti-disinformation and internet studies sector’s wagons however remain circled, increasingly so with the current blunt crack down on universities. The lesson of how we got to this point is being lost in this new struggle, but must be learned to avoid repeating the same mistakes.

With thanks to the liber-net team.

A previous version of this story wrongly stated the $979m Peraton award was not in a public database, however it can be seen on USAspending here.