How a Scientific Cartel Protects Fraudsters and Rakes in Billions of Taxpayer Dollars
Corrupt scientists rarely face accountability. The real victims are everyone else.
I was 22 when my grandmother forgot me.
It took her 12 years to die from Alzheimer's. It started with little things, like where her glasses were or what day it was. Soon she didn't know who I was. For a while, she addressed me as her son, but then, as the disease ate away more of her mind, she forgot him too. Then I was the young, handsome version of her husband, until he too faded away. After a while, I was just a nice young man who came to visit her.
The rest of the time, she was afraid: waking up in an unfamiliar world, surrounded by people she'd never met, confused that she wasn't back home in Minnesota, where she'd grown up. It hit my mom the hardest. She did everything she could to take care of her own mother, watching the brilliant, kind woman she knew rot into a husk of her former self.
My grandmother died on Christmas Eve. As sad as it was, it was a blessing for my mom, who was finally freed from her duty of watching the woman she loved the most waste away.
The Alzheimer's Researcher Who Became a Poster Child for Academic Fraud
Sylvain Lesné, a neuroscientist at the University of Minnesota, published a paper in Nature in 2006 claiming to identify a specific amyloid beta protein assembly as the direct cause of memory impairment in Alzheimer's. This reinvigorated the amyloid hypothesis at a moment when skepticism about it was ramping up. The National Institutes of Health (NIH) devoted $1.6 billion to projects that mention amyloids in 2022 alone, nearly half of all federal Alzheimer's funding that year. Lesné was a star.
But there were rumblings. Numerous amyloid drugs made it to trials with billions invested by pharmaceutical companies. They failed repeatedly. A question arose in the pharmaceutical community: How can this be right? How can the trials keep failing if the underlying research is correct?
In 2022, the Vanderbilt neuroscientist Matthew Schrag uncovered evidence that images in Lesné's paper had been manipulated. Science magazine found more than 20 suspect papers by Lesné, with over 70 instances of possible image tampering. Nature retracted the paper in June 2024. Every author except Lesné signed the retraction. Lesné himself resigned from his tenured position at the University of Minnesota on March 1, 2025, three years after his fraud was exposed.
More news and details trickled out over time. Charles Piller's 2025 book Doctored talks about the Amyloid Mafia, a nickname for a network that had prioritized novelty over replication and marginalized dissenters for decades. Anyone questioning the amyloid gospel was pushed out and watched their funding vanish.
When I first picked up that Science article, I hadn't considered academic fraud as something that was real and widespread. As I thought about it more, I was filled with a deep, bitter hatred. For his own pride, greed, and acclaim, this man had doomed millions of people like my grandmother to slow, horrible deaths and millions more like my mom to agonizing years as caregivers.
Lesné resigned, but was still rich. None of his grant money was clawed back. The system that was supposed to catch this—peer review, university compliance, journal editorial boards—failed repeatedly for years.
How Much Academic Literature Is Made Up?
Lesné was not a lone bad apple. The rot and corruption of academic research are systemic and structural. Daniele Fanelli's 2009 meta-analysis of survey data in PLOS One showed that approximately 2 percent of scientists self-reported fabrication or falsification—and 14 percent reported witnessing it in colleagues. Self-reports mark the floor, not the ceiling.
J.B. Carlisle's 2021 paper, "False individual patient data and zombie randomised controlled trials submitted to Anaesthesia," showed that out of 153 trials with individual patient data available, 44 percent had untrustworthy data and 26 percent were zombie trials animated entirely by false data. In a 2025 PNAS study, researchers estimated that the number of fraudulent publications is doubling every 1.5 years, while legitimate publications double every 15.
At least 400,000 papers published from 2000 through 2022 showed signs of coming from paper mills. Former BMJ editor Richard Smith asked, "Is it time to assume health research is fraudulent until proven otherwise?" A 2015 Lancet comment by Richard Horton put it bluntly: "Much of the scientific literature, perhaps half, may simply be untrue."
The Peer Review System Is Susceptible to Scientific Fraud
Mancur Olson's The Logic of Collective Action explains why systems that everyone knows are broken continue. When the benefits of cheating are concentrated in a small group but the costs are distributed across a vast population, the cheaters organize and the victims don't.
Lesné captured millions of dollars in grants, plus tenure, prestige, and conference invitations. But the millions of Alzheimer's patients and families, the wasted taxpayer money, and the honest researchers who could have gotten those grants instead—they all suffered for those benefits. None was hurt directly enough to trace culpability, but they suffered nonetheless.
Academia suffers from second-order Olson problems. Peer reviewers are unpaid volunteers with no upside for catching fraud and with significant social downsides for making accusations. University administrators have concentrated incentives to protect grant-winning faculty. The University of Minnesota investigation took roughly two and a half years and produced only a single resignation: Lesné's. There were no legal consequences, no clawbacks.
The peer review process helps enable this fraud. The economist Bruce Yandle called it the "Bootleggers and Baptists" phenomenon: A group with strongly held moral beliefs will end up working with people interested in exploiting it financially. Self-righteous gatekeepers say peer review is required for integrity. As a byproduct, the publishing oligopoly extracts billions in profit by charging for access to taxpayer-funded research. Paper mills have become a shadow market worth, by one conservative estimate cited by Nature, hundreds of millions of dollars per year by publishing any slop they get their hands on, knowing that researchers desperate to publish would rather cheat than starve and that sociopaths would happily buy authorship with a credit card.
An entire extractive industry that charges universities to access their own faculty's work, failing at its literal one job: integrity in publishing.
The system actively selects for corruption. The goals of individual scientists—getting published and getting grants—are structurally misaligned with the goals of science itself: the pursuit of truth. The Food and Drug Administration Amendments Act of 2007 legally required public posting of clinical trial results on ClinicalTrials.gov. A 2015 New England Journal of Medicine analysis showed only 13.4 percent compliance in reporting summary results within the required 12-month window. The government has the authority to fine violators up to $10,000 per day; it could have collected $25 billion according to a 2015 STAT investigation. But it collected essentially nothing, because the agency doesn't want to fight the powerful institutions it regulates.
When institutional gatekeepers have failed, the answer is not yet more gatekeeping by corrupt watchers and underhanded dealers. It's radical transparency and the free exchange of information.
Science must be a free market of ideas, but now it's a cartel. NIH grant funding is centrally planned science. A small committee directs billions, yet is structurally incapable of knowing which directions are most promising.
Luckily for us, information wants to be free.
AI Can Empower Independent Scientists To Fight Academic Corruption
Professionalization and credentialism dominated science in the 20th century, suppressing independent research. But Charles Darwin was an amateur naturalist, Benjamin Franklin a printer, Michael Faraday a bookbinder. I believe the world is not only ready but eager for a new wave of citizen scientists, independent researchers, and curious minds to hunt through those monumental piles of data, to burn through corrupt and deceitful journals, and to replicate the most audacious claims in pursuit of truth.
It isn't intelligence or motivation that limit independent research. The biggest limit is the labor cost of accessing, reading, cross-referencing, and synthesizing enormous bodies of literature. The advent of incredibly powerful AI has collapsed that cost by orders of magnitude. A motivated researcher with a ChatGPT Pro subscription can do in an hour what previously required months, a research team, and institutional library access.
The remaining limits are rigor, access to data, and the ability to conduct experiments. That's where open infrastructure comes in, and that's what I'm helping build today.
An acquaintance on X once told me that no one is translating Chinese research preprints to English. I smirked in disbelief. "There is literally no way that's true," I thought—and yet I looked, and it was. Thousands of Chinese preprints sat entirely untranslated with no real reference. A ripe opportunity to do something important was in front of me, and I decided to go and pick the low-hanging fruit.
ChinaRxiv.org now has 26,000 high-quality translated preprints, with the goal of continuing to expose information, to make it so researchers with high integrity have access to absolutely everything they can get their hands on to do their best work. If artificial superintelligence arrives soon, it probably won't particularly care about what language any information is in—and as we train bigger and stronger models, getting information in datasets that are easily scrapable and accessible is a big win.
From there, a new opportunity emerged: Soviet-era Russian academic literature, a vast corpus of pre-1980s papers that were never translated into English. Through conversations on social media, I met some like-minded researchers and connected with the Research Revival Fund, sponsored by the Analogue Group. Together, we are getting ready to translate large stretches of previously untranslated Soviet academic papers.
There are entire research traditions in material science, theoretical physics, mathematics, and biology that were conducted in Russian but never entered the English scientific conversation. It's still sitting there, untranslated, waiting to be found and introduced to the world.
The Real Victims of Corrupt Science, Slow Research, and Constrained Information
The victims of corrupt science and siloed information never know they were harmed. The child who dies because her cure was delayed by a few years never knows the hurt that she received. The researchers who didn't get the grant for their novel idea because the cartel was funding their buddies on the backs of fraudulent graphs never knew they got screwed. In light of all this disappointment, we have a choice: Continue trusting failed institutions that are mired in corrupt incentive structures, or build a parallel world that routes around these gatekeepers. Let free information and free markets do what centrally planned science cannot, and let the best works survive.
Real people suffer and die when science is slow, when science is corrupt, when science is constrained. When you remove barriers to information and let individuals act on dispersed knowledge, you get outcomes no central planner could ever dream of.