Last week, a 28-year-old lab worker died in Siberia after developing a severe pneumonia of unknown origin. Under ordinary circumstances, the death of a young woman from unexplained pneumonia wouldn’t attract much international attention. But Darya Shipilova worked at the Irkutsk Anti-Plague Research Institute of Siberia and the Far East, a facility that handles some of the world’s most dangerous pathogens, including Yersinia pestis, the bacterium that causes plague.
What happened next has done little to allay concern.
Reports circulated that Shipilova had contracted pneumonic plague following a lab accident involving a broken test tube. Russian authorities deny this. They say there is no evidence of an accident involving dangerous pathogens and that testing has not established a connection between her illness and her work. Yet around 200 contacts have reportedly been placed under medical observation, parts of the hospital where she was treated were quarantined, and Anna Popova, the head of Russia’s public-health agency Rospotrebnadzor, flew to Siberia for an emergency meeting. One local government official confirmed over the weekend that Shipilova had died from the plague, only to later edit his statement to add the word “possibly”. Meanwhile, the world has seen footage of men in hazmat suits stalking the streets of Irkutsk.
For the moment, there is no evidence of a spreading plague outbreak. No further infections have been identified among Shipilova’s contacts, and the World Health Organization (WHO) has assessed the public-health risk as “moderate to low” in Irkutsk, and “low” for the country as a whole, while seeking further information from Moscow.
But there is a more unsettling question lurking behind the headlines. What exactly goes on in a “plague institute” in Siberia? Might the organism involved have been altered in some way? Since Russia inherited part of the Soviet Union’s enormous biological warfare infrastructure, is it conceivable that this was not simply a laboratory accident, but evidence of something much more dangerous?
These are all legitimate questions. They are also questions for which, at present, we don’t have the evidence to answer.
So what do we know? It’s important to understand that there is nothing inherently sinister about having an anti-plague institute in Siberia. Plague did not disappear with the Black Death. Yersinia pestis still circulates naturally in populations of rodents and their fleas, including across parts of Russia, Mongolia and Central Asia. While human infections are relatively rare, the Irkutsk institute grew out of this natural prevalence. The regional anti-plague institute for Siberia and the Russian Far East has, for more than a century, monitored dangerous infections, investigated natural disease reservoirs and maintained specialist teams capable of responding to outbreaks of plague, anthrax, cholera and other serious infectious diseases.
In other words, its existence makes epidemiological sense. Nor is it uniquely Russian. Across the world, labs which are capable of working with dangerous pathogens are maintained. They are essential for diagnosing outbreaks, developing vaccines and treatments, and understanding diseases that might threaten us in future.
Russia does, however, have baggage. The Soviet Union secretly operated the largest-known biological weapons programme in history: Biopreparat. Soviet military scientists were working on pathogens which caused anthrax, smallpox and investigating whether they could be turned into effective weapons. Founded in the Twenties, it continued clandestinely even after they had become signatories of the 1972 Biological Weapons Convention. This history matters because it inevitably colours the interpretation of unexplained incidents at Russian biological facilities. Let’s not forget the smallpox outbreak of 1971.
But history is not evidence of what happened in Irkutsk.
There is currently still no evidence that Shipilova encountered an engineered strain of plague; still less that Russia is developing a “pandemic weapon” at the institute. Could Yersinia pestis be used as a biological weapon? Certainly. Its potential has been recognised for decades. Pneumonic plague is particularly frightening because of its historical associations. Infection of the lungs can progress extremely rapidly and can spread between people through respiratory particles. Untreated, it is frequently fatal.
But plague today is not the Black Death. It is a bacterial infection for which effective antibiotics are readily available. If recognised and treated quickly, patients have a good chance of recovery. Public-health authorities also know how to contain it: isolate suspected pneumonic cases, identify and monitor contacts, provide preventive antibiotics where necessary and protect healthcare workers.
This also puts the Russian response into perspective. The extensive quarantine measures have fuelled speculation that the authorities know something they are not saying. Perhaps they do. But the measures themselves are not evidence of an engineered pathogen, nor even evidence that Shipilova acquired her infection in the laboratory. If doctors genuinely suspected pneumonic plague in somebody who had already been treated in hospital, isolating contacts and healthcare workers is precisely the precautionary response one would expect.
The more disturbing hypothetical is not that this was ordinary plague, but that it was an engineered form of the pathogen, deliberately altered to make existing defences less effective. Modern biotechnology raises possibilities that 20th-century biological weapons programmes did not possess. Pathogens might, in principle, be altered in ways that change their disease-causing properties or reduce the effectiveness of existing medical countermeasures.
While that possibility deserves serious attention, it must not be confused with evidence that it happened here. Indeed, one of the great difficulties with biological threats is precisely that the same scientific capabilities can serve very different purposes. Scientists legitimately study how pathogens spread, why some variants cause more severe disease, and whether treatments will remain effective. But the knowledge and techniques needed to defend against biological threats can also potentially be repurposed to deliberately cause harm. The boundary between defensive research and activities that could contribute to weapons development is therefore not always visible simply by looking at the equipment inside a laboratory.
And this problem is becoming more important because the number of labs capable of conducting high-consequence biological research has grown. The Global BioLabs initiative, which I established at King’s College London, has documented a substantial expansion in maximum and high-containment laboratory capacity around the world. There are good reasons for this. Covid taught governments that the ability to detect emerging diseases, characterise pathogens and develop medical countermeasures must be part of national security.
Yet more labs and more work with dangerous pathogens also means more opportunity for something to go wrong. Of course buildings, protective equipment and elaborate operating procedures will reduce risk enormously. But they cannot abolish it. Equipment fails. Samples are mishandled. Procedures are breached. Humans tire. Researchers can be infected without immediately recognising their exposure.
We know this because it has happened before. After the 2003-SARS epidemic had been brought under control, laboratory-acquired infections subsequently occurred in Singapore, Taiwan and Beijing. In Beijing, one such infection led to further transmission outside the laboratory. But, accidents are not proof that dangerous-pathogen research should stop; they are proof that humans work in laboratories.
Covid should have forced us to confront the second half of this problem: not merely how to prevent lab accidents, but how to establish whether one has taken place.
Within weeks of Covid reaching Europe, when discussion of the origin of the enfolding pandemic was already becoming politically toxic, I argued that a credible investigation was needed to examine all plausible routes of emergence. Evidence of SARS-CoV-2‘s evolutionary history could not, by itself, establish how it first entered the human population. A naturally occurring virus could pass directly from animals to people. It could also, in principle, infect somebody collecting animals or samples in the field, or someone subsequently working with them. Distinguishing between those routes requires different kinds of evidence.
Instead, the Covid origins question became trapped in geopolitics. On one side, to raise the possibility of a research-related accident came to be treated as equivalent to promoting a conspiracy theory. On the other, the absence of transparency was taken as proof that a lab leak had occurred. Intelligence agencies were asked to resolve questions that would require scientific, epidemiological and forensic investigation. Years later, we still lack a definitive explanation that commands broad international confidence. That failure is directly relevant to Irkutsk.
Traditional outbreak investigations start with straightforward questions. Who became ill? When? Where? What did they have in common? What were their likely exposures?
But when a plausible connection to a laboratory emerges, investigators need to be able to ask another set of questions. What pathogens were being held there? Which strains? What work was being conducted? Who handled them? Were any incidents recorded? Did staff report unusual illnesses? What do clinical samples, sequencing data, laboratory records and occupational-health records show?
The purpose of such an investigation is not to prove that a lab caused an outbreak. It is to provide a credible way of determining whether it did or didn’t. That capability remains surprisingly weak internationally. WHO can investigate disease outbreaks and request information, but it cannot compel a sovereign government to provide unfettered access to outsiders. Other international protocols, including the Biological Weapons Convention and the UN Secretary General’s mechanism for investigating allegations of biological-weapons, are primarily concerned with suspected use of deliberately engineered pathogens. But they were not designed as routine forensic systems to answer a more basic question: did an outbreak occur naturally, or was it the result of a laboratory accident?
This creates a dangerous vacuum. We do not know whether Shipilova died from plague. We do not know whether she was exposed to anything at the Irkutsk institute. There is no evidence that the pathogen involved was engineered, and no evidence that this episode reveals a Russian pandemic-weapons programme. But nor should the world have to choose between taking Moscow’s assurances on trust and imagining the most frightening alternative.
What is needed now, then, is quite mundane: a clear diagnosis; an account of what Shipilova was working on; confirmation of whether a lab incident occurred; the results of testing and sequencing; and, ideally, sufficient independent scrutiny to make those findings credible. Essentially, we need informed transparency.
The gap between what is known and what is imaginable is where speculation flourishes. Covid demonstrated just how politically destructive that gap can become. The lesson from Irkutsk is, therefore, not that every unexplained illness in a biological researcher should be treated as evidence of a lab leak — still less a bioweapon. It is that, six years after Covid began, we still lack an international system capable of showing convincingly when one is not.