The Implications of the Lab-Leak Hypothesis

A cell infected with COVID-19.
An electron micrograph of an apoptotic cell infected with the SARS-CoV-2 virus. Photo: Image Point FR/NIH/NIAID/BSIP/Universal Images Group via Getty Images

Nothing has changed but the narrative. A majority of Americans now believe that the coronavirus emerged from a lab, not nature, and in recent weeks a new openness to the lab-leak theory has taken over “nearly all mainstream media,” as my colleague Jonathan Chait put it. But the material case for the hypothesis remains essentially unchanged from the version advanced by Nicholson Baker, in this magazine, in January — indeed more or less unchanged from the version that appeared in a September profile of Alina Chan in Boston magazine, or for that matter from the version outlined on Medium by Yuri Deigin last April, just a few months into the pandemic. Now, though, Nate Silver estimates the odds of a lab leak at 60 percent. The Washington Post has published five front-page stories about it in the last few weeks. Vanity Fair’s Katherine Eban published a long account of the lab-leak saga that treated the independent researchers first calling attention to the theory as heroic detectives. The Daily Mail sent a reporter and a photographer to stake out the Rockland County home of Peter Daszak, a once-obscure zoologist now under scrutiny for his role in global “gain of function” research to make viruses more dangerous.

Prominent scientists long on the sideline, like Scott Gottlieb and Peter Hotez, have begun echoing calls for a deeper investigation. Both the National Security Council and the Director of National Intelligence have made a point of emphasizing their agreement with the previous administration that the pandemic’s origins are a very open — and very important — question. “It might have started in the wild, or it might have started in a lab,” the scrupulous science journalist Daniel Engber wrote, as the flurry of reconsiderations settled onto the front pages of the country’s newspapers late last month. “We know enough to acknowledge that the second scenario is possible, and we should therefore act as though it’s true.”

Practically speaking, the media as a whole is already there, having moved in just a few weeks quite close to “acting as though it’s true” — a sign that the Trump era of American political epistemology may be mercifully receding, with liberal and center-left publications feeling now much freer to consider possibilities the president had once made functionally unthinkable. What comes next is not yet entirely clear, but one striking possibility raised by the public trajectory of the lab-leak narrative is that an increasing gravitational force tugging on any story will be its relevance for America’s growing rivalry with China — one of very few areas of broad agreement between Trumpworld and the D.C. Establishment foreign-policy “blob” that supplanted it in January.

In fact much of the new lab-leak “consensus” has been built by China hawks — Josh Rogin in the Washington Post, former deputy national security adviser Matthew Pottinger, former secretary of State Mike Pompeo, the Wall Street Journal and especially its editorial board — both inside and outside of government. In an eye-opening response to Eban’s Vanity Fair piece, which presented those within the Trump administration pushing for a further investigation as hitting a brick wall of institutional resistance, former assistant secretary of State Christopher Ford alleged Thursday that the lab-leak investigatory team had been conducting briefings both throughout the State Department and with “interagency partners” without even subjecting their central claims to review by scientific experts or the intelligence community. When he finally persuaded them to, even a panel of largely sympathetic experts found the evidence quite circumstantial and the aggressive lab-leak case built on it irresponsibly overstated. A lab-leak origin did seem possible, but a committed team of State Department insiders hadn’t been able to assemble much more evidence for it than Yuri Deigin or Alina Chan or Nicholson Baker had. The eventual result was that Pompeo downgraded his rhetoric and retailed a considerably more modest version of the hypothesis when he took it out on the road. According to Ford — who describes himself as very much open to the possibility of a lab-leak origin, just focused on making a well-substantiated case — Pompeo had originally wanted to declare publicly that it was “statistically impossible” for the disease to have come from anywhere but the Wuhan Institute in his effort to pin blame for the pandemic on China.

But the question of blame is a complicated one, even if one credits the lab-leak theory, since, as Baker documented, the work being done at the Wuhan Institute was in partnership with American scientists and institutions, and was funded in part by both the NIH and the Pentagon. To a degree that is hard to fathom given conventional wisdom about “the new Cold War,” the dangerous research at the heart of the lab-leak hypothesis was conducted largely in the spirit of basic cooperation and coordination, even though both countries regarded it as sensitive work bound up in national-security interests. This is a basic confusion of the whole “new Cold War” framework: The two most powerful countries in the world are, transparently, rivals, and yet they are also, in almost inextricable ways, partners. They are not — as the Cold War analogy suggests — competitive, self-contained empires operating from incompatible ideologies and separated by an Iron Curtain; they are something much more complicated and intertwined, if not quite one economy ruled by two governments. The midcentury U.S. did not import its prescription drugs from the Soviet Union, or the equivalent of its flatscreen TVs and iPhones, and if a Hollywood actor made an offhand remark about the refugees who fled the Russian revolution he wasn’t shamed into recanting — in Russian. (This sort of apology was what happened when the actor John Cena, promoting the ninth Fast and the Furious film, casually referred to Taiwan as its own country.)

Whether or not the coronavirus pandemic is the result of a lab leak, the Chinese government has behaved terribly in blocking any real investigation into its origins. And if the pandemic was the result of a leak, those who let it happen and those who presumably helped cover it up deserve some real blame, for the millions who’ve lost their lives and all of those who have otherwise suffered from the virus. But so would the American scientists who inaugurated this kind of research, and those who oversaw and helped fund it. Indeed, if we take the lab-leak theory seriously — even, as Engber suggests, “act as though it is true” — the most direct lesson is ultimately much simpler than geopolitics. It would mean that what is probably the gravest global public-health crisis in a century bore a human signature. That signature would spell “hubris,” since even the most innocent possible lab-leak origin story still involves the large-scale hunting, collecting, gathering, and transporting of exotic animal viruses for storage in centralized facilities — facilities often much closer to human populations than the diseases had been “naturally,” and which, in at least this one very consequential case, had failed the basic functions of security and safety. If Chernobyl canceled our nuclear future, or at least delayed it a generation or two, then surely a lab-leak version of COVID-19 might entirely eliminate this genre of virus research, too.

I spoke to Marc Lipsitch, the Harvard epidemiologist who has been for a decade now perhaps the most prominent advocate for at least reining in, and possibly ending, gain-of-function research — arguing all along, and well before this pandemic began, that the risks of accidental outbreaks could not be justified by the benefits the research could offer.

Over the last few weeks, the narrative around the lab-leak theory has really changed, at least as I see it. Does it look the same to you?
I think that last year the theory lay mostly dormant because everybody who had the relevant expertise was busy and was trying to focus on fighting the pandemic rather than investigating the origins.

What changed?
My idea of the narrative change is that the WHO report [which dismissed the lab-leak theory out of hand] came out. A lot of people were very frustrated by that.

Including, somewhat amazingly, the head of the WHO, who felt compelled to immediately announce that, in his view, all possible origins remained on the table.
And then it seems like our letter in Science [which called for a further investigation] sort of had the opposite effect of last year’s Lancet piece [which essentially dismissed the possibility of any non-zoonotic origins] — in the sense of what was once a live hypothesis had been killed, by branding it as like climate denial. We sort of brought it back.

The media has an amplifying role in both episodes, I think, though I also tend to think that some of the recent reconsideration has been a bit heavy on media criticism and a little less focused on the practical implications — what it would or should mean if the disease really did come out of a lab.
The media has understandably been really confused. I said to someone yesterday, I don’t really like groups of scientists standing on a pedestal and saying, we believe X, and you should believe X. I think it’s much better to say we believe in X because A,B, C, and D — that’s a very different thing.

But that evidence — that A, B, C, and D — am I mistaken in thinking they’re basically unchanged over the last few months? We don’t really know any more than we did before the narrative turned, right?
Yeah, yeah. I agree. It doesn’t seem like much more evidence has come out.

Do you expect we will get more? My own intuition would be that we’ll likely be living in this zone of uncertainty for quite a while, and possibly forever, in the sense that, barring the possibility of a Chinese whistleblower or the granting of access to the database that was taken down before the pandemic, we’re not likely to see clear evidence of a lab leak. And since we haven’t yet seen a plausible account of an intermediate zoonotic host, that’s probably not something we can count on either — a lot of people on the zoonotic side of the argument have been pointing out that we didn’t nail down the origin of SARS for more than a decade, but actually scientists had a quite good idea that civets were the intermediate host within months. Is that how you see it? Or do you think there will likely be some resolution? 
No, I think there’s a very good chance we will never know the answer. But I do think that as a society and as a country that is associated with some of this work, we have a responsibility to try and get access to medical records, to personnel records and serology. And even to know what the testing program was in the Wuhan Institute. But maybe that’s not going to happen because China doesn’t want to share it. But [the molecular biologist] Richard Ebright was on Faye Flam’s podcast and made an interesting point, which is that government progress reports from grants and grant applications and other materials could be the source of some information, and also papers that were submitted and maybe not all published. Those might give us some information, and I think there is some opportunity to dig into those. Maybe the intelligence agencies are already doing that. But I agree this is a very good chance that we’ll get no full resolution.

But I also think that many people would have been less critical if the WHO report had said here’s all the stuff we went through to look at the lab leak, and there’s no evidence — meaning there’s no record of anyone being sick and no record of anyone’s serology and no record of any virus that matches. But none of that negative evidence was documented.

So then how do we proceed? How should we? Presumably as part of this narrative shift, the American public and to some degree, the global public, is becoming at least more aware of this kind of research and its potential risks, and presumably at least to some degree, more concerned about it. How should we go about addressing those risks? As far as I understand the closest we’ve come in, the U.S. was this sort of pause on funding, in part a result of your advocacy during the Obama years, but that’s actually quite limited. It’s nothing like, say, the global nuclear nonproliferation system we have in place. So how would you like to see our protocols over this research evolve in response to this pandemic?
First of all, we have to define what we mean by this research. Because it’s not clear that the kind of research that my colleagues and I were focused on trying to call attention to had anything to do with this epidemic.

By which you mean gain-of-function research, right?
Such research was being conducted. There was a pandemic. But the lab hypothesis doesn’t require that it was an engineered virus.

It could’ve just been captured in the wild as is, stored in the lab untouched, and then gotten out.
Right. So I think they’re really two broad categories of research to talk about. One is getting viruses out of nature and trying to understand them, which nobody up till now has ever said we shouldn’t do. Some people have said we shouldn’t spend billions of dollars doing it.

Is that your position?
I think it’s not actually that productive a thing to do, but nobody has said that as a biosafety matter it should not be allowed, at least that I’m aware of.

The other category is manipulation of viruses to become more transmissible or more and more deadly or capable of immune escape. That’s the category that many of us were focused on in 2014 and 2015. I think both categories need some scrutiny, but I also think some of those kinds of research are unbelievably important and should be done, not banned, and indeed funded more, not less.

What do you mean? 
Well, one example would be a kind of study that was originally caught up in the moratorium, and then removed from it — adapting coronaviruses to mice, so we could study vaccines and drugs in mice.

They made the argument then, successfully, that first of all, this is really important for public health, which I agree with. And secondly, that it’s a much different kind of risk.

How so?
The goal of success in gain-of-transmission experiments is to make a new pandemic pathogen. Otherwise, you’ve failed. The goal of mouse adaptation studies is to make a mouse-adapted pathogen, which if anything is probably going to be less likely to transmit in humans — you’re taking a human pathogen and adapting it to mice. That’s an example of research I think should be done more, not less.

But while that might be safer, doesn’t it also offer less value in terms of vaccine development, compared to working on pathogens that were designed to be transmissible in humans?
No, I would say the opposite. The benefit of adapting a coronavirus to mice is that you can now study the mice, which you need to do to develop a human vaccine — this is an essential step in the drug-development process. You need some smaller animal that you can buy by the dozens to allow you to really study the virus. That’s really important work.

By contrast, the vaccine argument for gain-of-transmission experiments was never made compelling to me. They sort of said, well, if we know what’s going to happen in nature, because we do it in the lab, then we can predict it and we can maybe make vaccines against it. But you don’t need a human-adapted virus in order to make a vaccine against it. You need an animal model, but you do not need a human. So it was a very sort of, well, we need to know what’s out there because we know how dangerous it is, but there were all sorts of very strong, implausible, scientific assumptions. Like, you know, what’s true in this genetic background will also be true in the genetic background a year from now.

As a broad principle, then, it sounds like you’re saying that research that adapts known pathogens to mice to allow us to study the disease and develop vaccines and drugs is worthwhile, but research that adapts anything to humans is too dangerous. 
I think the even broader principle is that high-risk, low-reward work should not be done. Low-risk, high-reward should be done. And the off-diagonals are harder, right? High-risk, high-reward; low-risk, low-reward.

I thought it was interesting that you recently told Noah Feldman you would be in favor of gain-of-function research on COVID, because it was important to understand what variants were likely to emerge. That seems possibly like a high-risk, high-reward proposition, and you came down on “yes.” 
Well, I think in some ways it’s lower risk because it’s already out there, and there is already that selection pressure, with the virus in millions of people every day. But I’d be more focused on the reward side, and really pushing on that. Because every scientist who gets a grant has successfully convinced somebody that their science is important. We’re all very good at that, the successful ones. But the facile way in which we do that in most grant proposals is different from a really hard-nosed scrutiny. How many drugs am I going to get out of this? And is there a better way to get the drugs that’s not dangerous?

Are you talking about asking these questions globally? It seems like it’s one thing to implement academic standards here, within American universities, and maybe also to shape grant standards from places like the NIH. But that’s just the U.S. How significant is our funding and our research priorities in setting the rest of the world’s agenda?
I think it’s hard to say what the role of the U.S. is, but I am somewhat optimistic that it’s greater than some other people think. For better or worse, if you get published in Science, you’re a big shot. If you get published in Cell — technically that’s Dutch, I guess. And if you got published in Nature, which is British, you’re a big shot. If you get published anywhere else, you’re not as big a shot. And so the attitudes of Western scientists who’ve reviewed for those journals — they, like, really do matter. I don’t think China would be doing as much gain-of-function as it is if it wasn’t seen as prestigious by Western journals. So that’s one thing.

Another is one that Tom Inglesby and I published in mSphere, right before the pandemic broke out for — an outline for what we thought the process should be for the government. What we said was we need an open and transparent risk-benefit analysis of each potentially concerning experiment where the names of the people involved are given, where their views are summarized, and where explicit arguments about what’s the risk versus the benefit is laid out so that people can be incentivized to rigorously scrutinize them.

How does it work now?
Right now, it’s an anonymous group at Health and Human Services that just says “yes” or “no” and doesn’t give reasons. At the funding-approval level, that’s what we proposed, and then the journals could become secondary enforcers of that standard.

You mean by refusing to publish research that hadn’t complied with those standards.
Which means that even if you’re in a country that doesn’t require any of these things, you still have to do something like that, if you want to publish it in the big-time places.

That still strikes me as relatively small-bore, honestly. If we’re proceeding from the assumption that, at the very least, it’s possible that this pandemic is the result of a lab leak, that means monitoring research becomes monumentally more significant and salient to the average person than it might’ve seemed a few years ago. We could end this pandemic with 10 million people dead — actually, some estimates suggest we could be there already, and it could double from here. Which is a scale of problem, and risk, that suggests a much larger policy response, one on the scale of the quasi-legal, quasi-political nuclear nonproliferation system. 
That is possible. But I’m not a legal person, I don’t know really how to do that. I’m not against that approach, I just think there’s a long way to get aligned on these sorts of mechanisms before that.

I wrote a short piece with David Relman in Foreign Affairs, making the argument that gain-of-function with flu was like the easiest possible case, and that therefore we needed to get it right, because it was only going to get harder. One reason is that it’s expensive, but increasingly things will get cheap and you won’t need a lot of funding.

It’s an illuminating contrast with nukes, and the half-effective system we’ve built to protect against them: We essentially have an expectation globally that any nuclear activity should be done in public, and if it is hidden, the global community presumes it is nefarious and that the nation behind it is a bad actor. Gain-of-function research may be expensive compared to other scientific research, requiring funding that can be monitored, but it’s not at all expensive compared to developing a nuclear weapon. And so it would seem to me that the transparency imperative would be even stronger here — and probably something we should try to implement before everything gets dramatically cheaper. 
Yeah. But I think the scope is the tricky part. I don’t think I would recommend something like the way we treat smallpox, for instance, where only certain amounts of the virus can be used. Only certain people can do it. You have to get permission.

I could not be more of a layman on this, but transparency seems possibly as important as permission. My own intuition is that if there had been an expectation of true transparency in place before this pandemic, it would’ve been conspicuous that the Wuhan Institute took its database offline last fall, and it would’ve been a lot less acceptable that it was inaccessible at the onset of the pandemic and that it continues to be. Alternately, if we had been able to monitor that database throughout we might have felt much more confident about what was going on there. 

Maybe that itself wouldn’t have prevented a leak, if there was a leak, but it might’ve meant that the world saw much more clearly from the outset that a leak was a strong possibility. And if that had happened — if it was a lab leak and we knew it was from the start — then my own intuition, at least, is that the response would’ve been really different globally—there would’ve been a lot more immediate panic in Europe and the U.S., and a lot less complacency about the risk, because people would’ve just been more scared of a disease that came out of a test tube than one that came out of a bat.

Almost certainly, many would’ve died, it might’ve been calamitous. But my own guess is that it would’ve been considerably less calamitous.  
That is true.

An earlier version of this post incorrectly identified pangolins as the likely intermediate host animal for the SARS; it was civets.

The Implications of the Lab-Leak Hypothesis