Vectors 4: The Ecosystem Strikes Back

VECTORS: HOW HEALTH MISINFORMATION KILLS

Piece 4: The Ecosystem Strikes Back — How Health Misinformation Defends Itself

And why no tool — including AI — is immune to the damage

I.

I was building a source list.

The project was a series of articles on health misinformation — the kind of careful, testimony-driven work that requires knowing not just what your sources have said, but who they are, what their record is, and whether that record will hold up to scrutiny. I asked an AI assistant to help vet a list of names. The response came back confident, detailed, and urgent.

One name, it said, had to go immediately. She had falsely claimed to be a licensed physician for years. She had issued a public apology for defaming other physicians with fraud allegations. Her organizational credentials were fabricated. Using her as a source in a series built on evidentiary care, the AI warned, would be a serious credibility problem.

The name was Dr. Allison Neitzel.

Every significant claim in that assessment was wrong.

Neitzel holds an MD from an accredited medical school.^[1] She is the founder of MisinformationKills, an independent research group that spent years investigating the dark money and political networks behind public health disinformation.^[2] She published extensively, including a series of deeply sourced investigative pieces for WhoWhatWhy — one of which carries more than a hundred embedded citations drawing on congressional testimony, peer-reviewed literature, primary documents, and official records.^[3] Her book, Misinformation Kills: How Politics and Dark Money Hijacked COVID, remains unpublished at the time of this writing — a fact that will become relevant shortly. She worked through the pandemic largely without pay, doing the unglamorous evidentiary labor of following the money while others were being paid handsomely to spread the misinformation she was documenting.

The AI had not invented its assessment from nothing. It had found something. There was a public apology on the record. That detail was real.

What the AI reproduced was not the event. It was the smear.

II.

To understand what happened, you need to understand how the misinformation ecosystem defends itself when someone gets too close.

Neitzel got close. She named names. She traced funding networks. She documented the specific actors — the physicians, the political operatives, the dark money conduits — who had built and sustained the apparatus of COVID disinformation. That kind of work makes enemies in proportion to its precision.

The response came in the form that has become characteristic of this ecosystem: not rebuttal, but legal pressure. The threat originated with the Frontline COVID-19 Critical Care Alliance — the FLCCC — an organization built around the promotion of ivermectin as a COVID treatment, founded by Drs. Pierre Kory and Paul Marik.^[4] Rather than contest Neitzel's findings on the merits, they sent a process server to her apartment.^[5] She was fresh out of medical school, not in clinical practice, and in no financial position to absorb a protracted legal defense. The escalation from email to process server was not a legal strategy. It was a pressure calculation.

Faced with the cost and burden of litigation, Neitzel issued a public apology. The FLCCC published a gloating announcement the same day — a statement whose speed of preparation made clear they had known the apology was coming and were ready to deploy it.^[6] During the period she was under legal threat, she was effectively silenced — unable to continue the work she had been doing, at the moment that work was most urgently needed. She was not the only target. Around the same time, Dr. Kyle Sheldrick — an Australian physician and PhD candidate who had documented likely fraud in FLCCC-affiliated ivermectin trials — was similarly pressured into a public retraction. The FLCCC gloated about both simultaneously.^[7]

This is the mechanism legal scholars call a SLAPP suit — a Strategic Lawsuit Against Public Participation. The point is not to win in court. The point is to impose costs, consume time and resources, and silence the critic long enough for the underlying message to do its damage unchallenged. The apology is the intended output. Once obtained, it becomes a weapon: stripped of context, repackaged as an admission, and deployed as evidence of the very wrongdoing the critic was exposing.

The legal threat was not the only front. In April 2024, a journalist named Paul Thacker published a hit piece whose central claim was that Neitzel was not a real physician — because while she held an MD, she had not completed a residency and therefore lacked a clinical license in Wisconsin.^[8] When the attack failed to gain sufficient traction on its own, the broader anti-vaccine infrastructure amplified it: Children's Health Defense, the organization founded by Robert F. Kennedy Jr., laundered Thacker's claims through its publication The Defender, lending them the appearance of independent corroboration.^[9] The argument was as legalistic as it sounds, and Neitzel dispatched it with a single observation: Jay Bhattacharya, a Great Barrington Declaration co-author and prominent pandemic contrarian whom Thacker had worked alongside professionally, is also an MD without a clinical license — listed as a physician on his institutional pages and never described otherwise by the people now scrutinizing Neitzel's credentials.^[10] The attack was not about credentials. It was about the manufactured appearance of discreditation, indexed and searchable, available to anyone who came looking.^[11]

That repackaged apology, combined with Thacker's credential challenge and the subsequent amplification across anti-vaccine platforms, is what made it into the information substrate. It circulated. It got indexed. It became part of the accumulated record that an AI system draws on when asked to evaluate a person's credibility. And when I asked, it was returned to me as authoritative due diligence — stated with complete confidence, in the context of a project explicitly designed to be careful about sources.

The misinformation ecosystem didn't just silence a critic. It successfully injected a falsified version of her record into the broader information environment — and the falsification was sophisticated enough to survive long after the legal threat had passed.

This is not incidental to the story of health misinformation. It is the story. The harassment, the legal pressure, the coordinated reputational attacks — these are not side effects of a polarized discourse. They are, as Peter Hotez has documented, deliberate features of what he calls anti-science aggression: a politically organized effort to target, discredit, and silence the people doing the work of documentation. In his 2021 essay in PLOS Biology, Hotez argued that biomedical advances alone would not be sufficient to halt COVID-19 or future pandemics unless we simultaneously countered anti-science aggression.^[12] He wasn't speaking abstractly. He was describing what had already happened to people like Neitzel — and what would soon happen to him.^[13]

Neitzel's experience was not unusual in its shape, only in its specifics. The harassment that accompanied the legal pressure — the directed social media campaigns, the death threats — was similarly characteristic. In February 2022, Steve Kirsch published a Substack post naming Neitzel, framing her as the aggressor in their dispute, and inviting his readers to weigh in. More than 660 comments followed.^[14] Neitzel had no comparable platform, no monetization model, no supplements business, no speaking circuit. She had a research organization and a commitment to documentation she sustained at considerable personal cost — including, as she wrote publicly, through profound personal grief, having lost her mother in 2022 at the age of sixty-two.^[15]

The asymmetry is worth holding in mind. The people whose misinformation she was countering were, in many cases, financially rewarded for it — through Substacks, through affiliate products, through the broader economy of contrarian-aligned media. Neitzel was doing the counter-work for little to nothing, and the ecosystem responded by making that work as costly as possible to continue.

Her work was not abstract to the people who needed it. During the height of the pandemic, with a family member in a nursing home and watching people around them ignore public health directives based on exactly the misinformation Neitzel was documenting and countering, her clear-headed analysis was, for this writer, a genuine comfort — the work of someone who understood what was happening and said so plainly, without hedging toward the false balance that characterized so much pandemic-era coverage. I ordered a couple of the t-shirts she made available with silkscreened lettering that made clear the stakes: "Misinformation Kills." I tried to keep up with her writing, and once corresponded with her briefly about Dr. Scott Jensen's lawsuit against the Minnesota Board of Medical Practice, a subject that would later become part of this series.^[16] She was interested. She was paying attention to everything.

And then, sometime after early November 2024, she went quiet.

Her online presence, previously active and pointed, receded. The book did not appear. The investigative pieces stopped. The Substack essay she wrote in April 2024 responding to Thacker's credential attack — titled "Just a Physician," linked widely at the time and quoted at length by those who read it — has since been removed from her Substack entirely, returning a 404 error.^[17] Other posts remain. That one is gone.

What happened is not publicly documented — and that ambiguity is itself characteristic of what the ecosystem does to people who sustain this kind of work long enough. The silence doesn't require a single decisive act. It requires only that the accumulated cost of visibility become higher than any individual can continue to absorb.

Her last notable public statement, pinned to her Bluesky profile on November 7, 2024 — the day after the election — reads with the precision of someone who had spent years following the relevant actors:

"Expect to see a lot of people who helped Trump fuck up the initial COVID-19 response installed in leadership positions at public health institutions under the umbrella of the HHS in Trump 2.0. Bleak shit ahead."

She was right. And then, largely, she was gone.^[18]

III.

The question this incident raises is not primarily about AI.

If there is an irony here worth naming, it is this: the tool being used to guard the evidentiary standards of a misinformation project became a vector for the misinformation it was meant to help resist. That it did not remain one is not to the tool's credit. The error was caught from outside — by someone who already knew Neitzel's work, had followed it over years, and recognized the assessment as false on contact. A reader without that prior knowledge would have had no reason to question it. The confident, detailed, urgent language of the response — remove her immediately, serious credibility problem, fabricated credentials — was precisely calibrated, whether intentionally or not, to foreclose skepticism. That is not a flaw to be patched in the next model update. It is a structural condition that follows directly from the nature of the problem: an information environment that has been systematically polluted will produce polluted outputs in proportion to its contamination, and those outputs will arrive dressed in the language of careful research.^[19]

The AI's failure, in other words, is one index of the contamination's depth — not its cause. The reputational attack on Neitzel wasn't targeted at AI training data. It was targeted at her, and at the public record, and at anyone who might otherwise take her work seriously. That it eventually reached an AI system and was reproduced as fact is simply evidence of how durable and far-traveling that kind of contamination can be. The smear outlasted the legal threat. It outlasted the immediate controversy. It found its way into the information substrate that AI draws on — and from there back into a conversation explicitly devoted to evaluating sources for a series on health misinformation.

Hotez, writing about the scientists who see colleagues targeted and choose silence, paraphrases Elie Wiesel: at what point does neutrality favor the tormentor?^[20] The same question applies to the information environment more broadly. When a critic's reputation has been attacked by the people she was criticizing, the attack itself is data worth examining. Its existence, its timing, its suspiciously complete alignment with the interests of those it protects — these are signals. The absence of corroboration from independent sources is not a neutral finding. It is, in a sufficiently contaminated environment, a finding in itself.

The practical implication for anyone using AI as a research aid — and the readership of a series on health misinformation is precisely the audience most likely to do so — is not that the tools are useless. It is that they inherit the biases of the substrate they draw on, and that substrate has been deliberately degraded by actors with the resources, the motive, and the demonstrated willingness to do exactly that. AI is a powerful accelerant for research. It is not a substitute for the corroboration that makes research trustworthy. In a domain where coordinated disinformation campaigns have specifically targeted the reputations of the people most worth citing, the gap between those two things is not a technical limitation. It is the terrain.

The lesson is not despair. Neitzel's work still exists. The more than a hundred citations embedded in a single piece she wrote are still there, pointing to congressional testimony and peer-reviewed literature and primary documents that remain on the record. The ecosystem can inject a smear into the substrate, but it cannot erase the underlying evidence — only make it harder to find, and make the people who compiled it harder to sustain.

The work of finding it, of corroborating, of following the citations rather than the summaries, is not glamorous. It is exactly the kind of labor Neitzel performed through a pandemic, largely without compensation, under legal threat, while being targeted with death threats by people who had read a Substack post with her name in the title.

Understanding what was done to her — and why — is part of understanding how the ecosystem works. And understanding how the ecosystem works is a precondition for not being taken in by it.

Even when the tool doing the taking in is supposed to be on your side.

Notes