In the early days of the pandemic, as billions of dollars poured into the hunt for novel treatments and vaccines, veteran Silicon Valley entrepreneur Steve Kirsch did what he’s always done: He went looking for an underdog.
Since making a fortune as the founder of Infoseek, an early search engine that was the Google of its day, Kirsch has spent tens of millions of dollars fighting humanity’s biggest threats. He prefers iconoclastic approaches, whether by directly funding asteroid detection or advocating for nuclear power to combat global warming.
With little government funding available for such work, Kirsch founded the Covid-19 Early Treatment Fund (CETF), putting in $1 million of his own money and bringing in donations from Silicon Valley luminaries: the CETF website lists the foundations of Marc Benioff and Elon Musk as donors. Over the last 18 months, the fund has granted at least $4.5 million to researchers testing the covid-fighting powers of drugs that are already FDA-approved for other diseases.
That work has yielded one promising candidate, the antidepressant fluvoxamine; other CETF-funded efforts have been less successful. But that’s not a surprise, according to researchers who conducted them: the vast majority of trials for any drug end in failure.
What has alarmed many of the scientists associated with CETF, though, are Kirsch’s reactions to the work he’s funded—both successes and failures. He’s refused to accept the results of a hydroxychloroquine trial that showed the drug had no value in treating covid, for instance, instead blaming investigators for poor study design and statistical errors.
He’s also publicly railed against what he claims is a campaign against drugs like fluvoxamine and ivermectin. And, according to three members of CETF’s scientific advisory board, he put pressure on them to promote fluvoxamine for clinical use without conclusive data that it worked for covid.
More recently, he’s adopted extremist positions on covid vaccines, which he alleges are “toxic.” He has claimed that one in 1,000 people who have received mRNA vaccines have died as a result, and even claimed the vaccines “kill more people than they save” at an FDA public forum, which was first reported by the Daily Beast.
As Kirsch has gone deeper into the anti-vaccine scene, many professional associates have increasingly distanced themselves from him. In May, all 12 members of CETF’s scientific advisory board resigned, citing his alarming dangerous claims and erratic behavior. Over the summer, the conflict reached his most recent startup, M10. Its board told him that if he wanted to remain part of the company he would have to stop making public anti-vaccine statements. In September, he resigned as CEO and gave up his board seat.
So how did a man once intent on furthering science become a source of misinformation that undermines the very research he funded?
Kirsch did a lot of things right when he set up CETF. To vet proposals, he recruited a powerhouse advisory board of prominent biologists, drug developers, and clinical researchers, led by world-renowned drug researcher Robert Siliciano of Johns Hopkins. (The fund borrows its nonprofit status from the 501(c)(3) Rockefeller Philanthropy Advisors, which managed its money until it quit, according to the Daily Beast; neither organization is related to the Rockefeller Foundation, which supports Technology Review's reporting on covid.)
While Kirsch had the final say in who received grants, no one I spoke with expressed concerns about what projects had been funded, or why.
“I agreed to do it partially because I respect Bob [Siliciano] so much, and partially because I thought the concept was excellent,” said former board member Doug Richman, a prominent HIV drug researcher at the University of California San Diego and former member of the fund’s scientific advisory board. “I think we did rigorous reviews of proposals for research.”
One of the first CETF grants was to investigate the antimalarial hydroxychloroquine. David Boulware, a researcher at the University of Minnesota, received $125,000 to test the drug against covid. The results would, eventually, set Kirsch on a collision course with the scientific establishment.
“Steve Kirsch was extremely helpful early on in the pandemic, stepping up to fund early treatment trials when the US government would not fund such studies,” Boulware told me in an email.
The drug was widely prescribed as a covid treatment for much of 2020, based on anecdotes and flawed studies. Boulware’s trial was part of a wider movement to bulk up the evidence base underlying standard covid treatments, and one of several trials that found no benefit to using hydroxychloroquine.
But the confusion provided a fertile breeding ground for skeptics. Indeed, some of the most prominent people spreading misinformation about ivermectin and vaccines today began by promoting hydroxychloroquine—including by claiming to “debunk” Boulware’s data analysis.
Kirsch, despite having direct access to the actual trial runner, eventually became convinced a “correct” interpretation of the data would show that hydroxychloroquine worked.
Boulware disputes that, and says that although Kirsch’s funding was important, his statements about drugs and vaccines have proven problematic. “I disagree with his interpretation of the data regarding several medicines and strongly disagree with his anti-vaccine nonsense,” Boulware wrote to me.
While he declined a phone interview, Boulware was recently the subject of a Mother Jones article about the harassment he’s received for his research on hydroxychloroquine and ivermectin.
A few months ago, Kirsch suddenly stopped promoting hydroxychloroquine—even scrubbing it from the CETF’s official list of trials it has funded. He wrote on his personal website that he’d been advised that being associated with the drug “would immediately trash my credibility.”
A more promising candidate
Another CETF grant, though, yielded far more exciting results. Drug researchers at Washington University in St. Louis reached out to Kirsch looking for $67,000 to finish a very small—but placebo-controlled—trial. They were giving covid patients the antidepressant fluvoxamine as soon as possible after diagnosis, based on anecdotes about the drug limiting the runaway immune response that causes many severe symptoms.
In October, the group reported that, while a few patients in the placebo group ended up in the hospital, none of the patients receiving fluvoxamine got sick enough to go.
In November, CETF gave the group an additional $500,000 for a phase 3 clinical trial that might show conclusive proof of efficacy. That trial has now been completed, and the researchers are analyzing their data. Several other trials around the world are in the final stages, too.
But the whole process has gone too slowly for Kirsch. Immediately after the results of the first fluvoxamine trial were released—but before they were published in a peer-reviewed journal—he wrote a post on Medium.com called “The Fast, Easy, Safe, Simple, Low-Cost Solution to COVID That Works 100% of the Time That Nobody Wants to Talk About.”
Medium banned him for misinformation. “Medium revoked my account for life. My crime? Telling the truth,” he tweeted. “Be warned!”
Since then, he has continued to promote fluvoxamine, along with ivermectin and hydroxychloroquine. When I asked him why so many experts in the field disagreed with him, he alleged there were efforts—either malicious or negligent—to suppress evidence of cheap, effective covid treatments.
“[NIH] doesn’t want any of these treatments. That’s why they didn’t even fund the fluvoxamine trial,” he told me. Instead, the government prefers to fund and promote new, proprietary drugs and vaccines, he says.
Reached by email, the two fluvoxamine investigators denied that there was any effort to suppress their research, and they were cautiously optimistic about their continued study. (One of them, Eric Lenze, was in fact giving a presentation on fluvoxamine to the National Institutes of Health the next day.) Both of them encouraged anyone reading this article to get vaccinated.
“Although there is evidence that fluvoxamine can prevent clinical worsening and the need for hospitalizations in outpatients with early covid-19, I have seen no good evidence that fluvoxamine is useful as a substitute for the vaccines,” co-investigator Angela Reiersen wrote to me.
“I got tired of arguing”
Kirsch is a serial entrepreneur who has spent decades pitching the next big thing, whether optical mice (Mouse Systems), document processing (FrameMaker), search engines (Infoseek), digital security (OneID), or e-commerce (Propel Software). His latest startup, M10, is a spin-off of a spin-off that sells a blockchain for banks. He has made millions from these projects, even if they have not turned him into a household name.
“You see this with people who have a lot of money, who think that reflects their intelligence,” Richman told me. “He considers himself an expert in something that he doesn’t have training or experience in, and he’s not following scientific methods to assess data.”
As trial results rolled in, that mismatch began to put a strain on Kirsch’s relationship with the fund’s advisory board. Several former members told me he began relentlessly pressuring them to promote the drug in media stories, often during exhausting, circuitous conversations.
“After one or two conversations like that, I got tired of arguing, so I started avoiding his calls.”Judith Feinberg, West Virginia University and former member of CETF's scientific advisory board
“Steve wanted to say, ‘Look, I’ve got all these famous [infectious disease] docs and researchers, and they all say give fluvoxamine a chance,” Judith Feinberg, one of the former CETF advisory board members and vice chair of research at the West Virginia University School of Medicine, told me. She understands complex, politicized pandemics—she was one of the first clinicians to specialize in HIV/AIDS, and she sat on the FDA advisory panel that approved the first antiretroviral drug.
But even she was drained by Kirsch’s constant attempts to override the data. “After one or two conversations like that, I got tired of arguing, so I started avoiding his calls,” she said.
To scientists, giving fluvoxamine a chance means running a large trial—not giving it to individual patients in the clinic, off-label and outside the context of active data collection and analysis. The board members I spoke to say they refused to publicly promote any drugs for off-label use and tried to explain to Kirsch that it’s incredibly common for exciting results from small trials to disappear in larger ones.
Peter Meinke, another former board member, spent nearly three decades in drug discovery at Merck.
“It’s really, really common for a small effect, something that looks exciting, to be a statistical fluke when you look at a larger population. It’s sad, but it’s true,” he told me. “With covid, 80% of your patient population does just peachy with no treatment at all, just a little bed rest and fluid. It’s actually much harder to parse out a signal than if you’re treating diabetes or cancer.”
In addition to the issues with fluvoxamine, advisors grew increasingly uncomfortable with Kirsch’s posts about ivermectin, which he has repeatedly claimed in blog posts and appearances in alternative media can be used together with fluvoxamine to prevent 100% of covid-19 deaths. (“The ivermectin data are trash,” Feinberg told me. “There’s nothing there.”)
Things took a final and dramatic turn once Kirsch started claiming the government was covering up vaccine deaths.
At the end of May this year, Siliciano emailed the other advisors to say that Kirsch had gone off the deep end and he was cutting ties. The rest of the board soon followed. (Siliciano did not respond to requests for comment for this article.)
Up for debate
Talking to Kirsch is an exhausting experience. He is frequently brash and interruptive, peppering dire warnings about vaccines with veiled aspersions toward Anthony Fauci and vague references to influential people who agree with him in private but cannot speak publicly.
In three phone conversations, as well as dozens of emails, his responses to questions about claims in this story were imprecise or constantly changing. He told me that while he and his family got vaccinated as soon as they were eligible, he got the idea that vaccines are dangerous from a man he hired to clean his carpets, who got very sick after receiving the vaccine. Elsewhere he has said he began questioning vaccine safety after an unnamed Twitter follower told him several family members died after getting their shots. He’s also recently increased the number of Americans he claim have been killed by the vaccine from 25,0000, to 150,000, or even “as many as 250,000” Americans.
Jeffrey Morris, director of biostatistics at the University of Pennsylvania Perelman School of Medicine, has made debunking Kirsch’s claims something of a hobby. On his blog, Covid-19 Data Science, he has extensively—and mercilessly—unpacked Kirsch’s “evidence” for the vaccine death claims.
In a recent post, discussing claims Kirsch made during a three-minute comment at an FDA public forum, Morris wrote:
“In spite of many pages of writing and claims of over a dozen ‘independent analyses’ verifying their results, their evidence falls far short of substantiating these dramatic conclusions, including a claim that vaccines have caused >250K excess deaths in the USA.”
And while Morris believes that all claims about vaccine safety should be properly vetted—“Is it possible there’s another rare side effect of the vaccines that we haven’t figured out yet? Yeah, it’s possible,” he told me—he also says that he has regularly seen Kirsch manipulate evidence so that it seems to support claims that are, in reality, baseless. In fact, he was unwittingly the source for one of Kirsch’s figures.
In September, Kirsch emailed Morris asking him to estimate the maximum number of deaths caused by vaccines. “Who knows,” Morris replied. “But not 150K. And not zero.”
Kirsch immediately forwarded the exchange to me and, I suspect, other journalists. “BOMBSHELL: Top biostats professor admits we have NO CLUE # of people KILLED by COVID vaccines,” he wrote. “He thinks # killed by vax could be anywhere between 0 and 150K people dead.”
Those who know Kirsch say this is a typical tactic. He’s adept at debate, rapidly shifting the premise of a conversation to put the other person on the back foot.
“He may not be a good scientist, but he’s smart,” says WVU’s Feinberg. “He’s very convincing. He might be a good snake oil salesman.”
I experienced this myself when, on one call, we discussed several studies. Kirsch told me that “meta-analyses are a higher level of evidence than randomized controlled trials.” When I responded that meta-analyses are only as good as the data they are based on, he said “I’d like to understand your source on that, because I can’t find a source that says a phase 3 trial is greater evidence than a meta-analysis.”
“When you characterize me, you need to say that Steve Kirsch doesn't go with majority votes on interpreting data.”Steve Kirsch
While combining the results of several well-designed trials can strengthen an argument or unearth patterns unseen in smaller samples, a meta-analysis is just the sum of its parts; any single well-done experiment is more useful than combining the results of several poorly done ones. Still, in the moment, his question threw me, and I stuttered.
Perhaps Kirsch’s most effective tactic, though, is simply his willingness to outlast everyone else. During our first conversation, which turned into a multi-hour Zoom session, Kirsch paced through the rooms of his cavernous house with his phone held at chest level, rarely looking down at the camera. Thirty minutes past the end of our scheduled time, he dropped his phone in the cupholder of his Tesla so that he could keep talking while he ran an errand.
“When you need to characterize me, you need to say that Steve Kirsch doesn’t go with majority votes on interpreting data,” he told me when I asked about his views on ivermectin, which he insists is a silver bullet against covid. “If you wanna find someone to debate me for ten thousand dollars, or a thousand dollars, I’m happy to do that, just for your benefit.”
Eventually, a press representative who was listening in, David Satterfield, unmuted his microphone to suggest we finish our conversation by email. After I ended the Zoom meeting, Satterfield called me to apologize for cutting us off. “I was just getting tired,” he said, before asking to speak off the record.
A web of influence
None of this would really matter if Kirsch’s views on vaccinations were private, or shared with a limited audience. But as Kirsch has clashed with the experts he initially surrounded himself with, he’s grown increasingly close to others who share his perspectives on vaccines—who have, in turn, provided a large and receptive audience to his claims about a fluvoxamine conspiracy.
His appearance on an episode of anti-covid-vaccine, pro-ivermectin pundit Bret Weinstein’s DarkHorse podcast, alongside Robert Malone, a prominent source of vaccine misinformation, introduced Kirsch to followers of the “intellectual dark web,” who have since embraced him as a fellow truth-teller. He’s also made several videos and podcasts with Vladimir Zelenko, the conspiracy theorist doctor who convinced Trump to take hydroxychloroquine.
While YouTube has repeatedly taken down the full video of the DarkHorse episode, various clips have been watched over 4 million times, and the full audio remains available on Spotify.
“The claim that ‘the spike is toxic,’ that came directly from the [DarkHorse episode]. I see it all the time on social media,” Morris told me. He’s probably the closest thing Kirsch has to a nemesis, regularly disputing his assertions in blog posts and private email exchanges with Kirsch and his friends. “I didn’t intend to spend a lot of time on Steve in particular, but that video was so influential.”
In June, after CETF’s advisory board resigned, Kirsch did a Facebook Live video with Zelenko and celebrity rehab coach Dr. Drew. In it, he claimed mRNA vaccines kill one in 5,000 recipients and dramatically increase the rate of miscarriages.
There is absolutely no evidence that either one of these claims is true, as Morris has carefully documented. Kirsch, though, often relies on the heartstrings to smooth over a lack of data. On Dr. Drew, he told a story about “a friend’s daughter” who had to get an abortion because of damage caused by the shot.
“The baby’s brain was split in half, and it was just covered with blood. It was so bad you couldn’t even see the baby’s body through all the blood,” Kirsch said. “They immediately ruled out the vaccine, because the vaccine is, quote, ‘safe.’”
Soon after his appearance on the DarkHorse podcast, several partners of his most recent startup, M10, expressed concerns about the increasing extremism of Kirsch’s vaccine views. “We asked Steve to tone it down. It was not compatible with his position as CEO to continue taking a very public stance on the vaccines,” Richard Char, M10’s general counsel, told me.
“He is very smart, and knows that he is very smart, and sometimes he behaves like he thinks he's the smartest guy in the room, whether he is or isn't.”Richard Char, M10
Kirsch’s response was to take his name off articles he’d written about vaccine deaths, changing the authorship to “VaccineTruth.”
On July 1, he tweeted from his personal account, “My publicly shared concerns regarding the safety of the COVID-19 vaccines may have had a negative impact on my company, M10. To protect M10 from my COVID-19 vaccination opinions, I will no longer post about my vaccination concerns here.”
He started a new pseudonymous account, @VaccineTruth2, to continue broadcasting messages. But even that didn’t last long. By the beginning of September, he was no longer the company’s CEO, replaced by his co-founder, Marten Nelson. He immediately tweeted an offer to give anyone $1 million if they could win a debate with him about vaccine deaths.
“He felt like he in good conscience had to speak out about covid, and so he made the decision to separate himself from M10,” says Char, who has known Kirsch since the 1980s.
The incident, he added, was “completely in keeping with his personality.”
“He is very smart, and he knows that he is very smart, and he—sometimes he behaves like he thinks he’s the smartest guy in the room, whether he is or isn’t,” he told me.
Saving the world has been a theme of Kirsch’s life for years. “There are two ways I’ve discovered that I may be able to save the world,” he told an IEEE Spectrum reporter in 2000. “One is to reduce the threat of nuclear war. Another is to identify an asteroid that is going to hit the planet.”
His efforts became more focused on medical research when, in 2007, he was diagnosed with a rare blood cancer. His foundation shifted focus to one goal—curing Steve Kirsch—by supporting one of the few scientists looking at the disease. After several failed attempts to stop the progression of his disease, he designed his own protocol for chemotherapy and doctor-shopped to find an oncologist who would give it to him. He’s now outlived his initial prognosis by several years.
“He’s a genuinely good guy. I mean, he really, truly has a heart of gold,” Char told me. “He’s spending his own money to do what he thinks is right. It’s motivated out of his sense of keeping people safe and advancing health care.”
But Kirsch is also motivated by an unsatisfied competitive streak. In that same IEEE Spectrum story about his then-new startup, Propel Software, he said he felt successful, but not famous.
“Mouse Systems is not a household word,” he told the journalist. “We didn’t come up with better mouse technology than Microsoft did. Infoseek lost out to Yahoo; it had a chance to grow bigger, but it didn't. And FrameMaker is still a niche product. Yes, these were successes, but the successes could have been bigger if we had really paid attention to marketing. I’m not going to make the same mistake again.”
“Now we’ve lost the high ground”
It is not unusual to be wary of developing science, or wrong to be skeptical of pharmaceutical companies. These huge businesses do often prioritize profits over human health: in 2009, Pfizer paid a $2.3 billion settlement over kickbacks and fraudulent marketing, including a $1.3 billion felony fine.
In 2013, Johnson & Johnson paid $2.2 billion for its own kickback and fraud scandal, including a specific $400 million fine for its subsidiary Janssen, which manufactures the covid vaccine. The US government accused Janssen of improperly promoting the antipsychotic drug Risperdal to dementia patients despite the drug increasing deaths in the elderly. The man who ran Risperdal sales, Alex Gorsky, is now CEO of Johnson & Johnson.
As a health care journalist, I started off firmly in the wait-and-see camp on mRNA vaccines. Thanks to the volumes of data and information provided by pharmaceutical companies and regulators, as well as large numbers of trials from independently funded research groups around the world, I now trust that they’re safe for the vast majority of adults.
I also think it makes a lot of sense to look for pre-existing drugs that can help treat covid symptoms. Over the next few years, millions of unvaccinated people are going to get covid; it’s vital to try to mitigate their suffering, as well as lessen pressure on the health care system.
But the best way to help people is through rigorous trials that show what drugs help which people, and at what doses and times—not by basing entire protocols on incredibly limited evidence.
Unfortunately, as Jeffrey Morris at UPenn points out, public health officials and scientists have done plenty to undermine their own authority, like claiming masks don’t work, downplaying the natural immunity conveyed by previous covid infections, and not doing enough public communication about vaccine safety surveillance systems.
”We don’t want to feed the anti-vaccine trolls, so we actively suppress clear scientific data. Now we’ve lost the high ground,” Morris told me.
And that is what has allowed Kirsch, and people like him, to become so influential. It’s a cycle that feeds mistrust and boosts the profiles of influencers who present themselves in opposition to official authorities.
“The collateral damage is that, now, a lot of people don’t trust scientific leaders or the scientific community. They’re finding alternative leaders to follow,” Morris said. “That’s what creates some of these heroes.”
This story is part of the Pandemic Technology Project, supported by The Rockefeller Foundation.
This new data poisoning tool lets artists fight back against generative AI
The tool, called Nightshade, messes up training data in ways that could cause serious damage to image-generating AI models.
Rogue superintelligence and merging with machines: Inside the mind of OpenAI’s chief scientist
An exclusive conversation with Ilya Sutskever on his fears for the future of AI and why they’ve made him change the focus of his life’s work.
The Biggest Questions: What is death?
New neuroscience is challenging our understanding of the dying process—bringing opportunities for the living.
Data analytics reveal real business value
Sophisticated analytics tools mine insights from data, optimizing operational processes across the enterprise.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.