Pandemics may arise from nature, and we are currently concerned about unprecedented spread of H5N1 avian influenza, with hot spots in Europe and the Americas. In the United States, widespread epidemics in dairy farms have affected the commercial milk supply. If the virus mutates to be easily transmitted between humans, a pandemic may arise. The other virus with pandemic potential is Clade I mpox, which has begun to spread to other countries from the Democratic Republic of Congo, prompting the WHO to declare a Public Health Emergency of International Concern.
We do not hear as much about lab-created pandemics, which may arise as a result of human error or terror. Lab leaks are exceedingly common in legitimate, regulated laboratories as catalogued by the American Biosafety Association. If such an accident involves a highly transmissible pathogen, a pandemic may arise. Lab-created pandemics can arise from illegal labs. The most notable discovery of a clandestine lab was on the ranch of the Rajneesh cult in Oregon, who perpetrated the largest documented bioterrorist attack in the US. The lab was discovered incidentally, almost a year after the attack. At the time of the epidemic, the CDC and local health authorities disregarded the suspicions of community members and a local politician, instead attributing the outbreak of salmonella to poor hygiene. Six months later, Rajneesh even confessed that his cult had done it, but was not believed. Yet despite so many red flags and warnings, the finding of a clandestine lab surprised everyone concerned.
Tracking biolabs
So, where are all the labs in the world, and how many do we track? For legitimate labs at the highest level of biosafety, BSL4, the number of labs is relatively low, and can be tracked more readily. For the next level, BSL3, which includes labs doing research on SARS-COV-2, HIV and influenza, there is no comprehensive global register. Such labs are in government, pharmaceutical, university and research organisations. The BSL2 labs are usually focused on high throughput diagnostics, such as private pathology labs. Pathogens by biosafety level varies by country, and even within countries, work on level 3 pathogens may occur in level 2 labs. Groups like Global Biolabs and DRASTIC have begun to track BSL4 and BSL3 labs, but comprehensive global registers are still lacking.
DIY bio and waste streams
The explosion of the Maker Movement and DIY labs is a global phenomenon, and there are many public DYI biolabs operating at BSL 1 level (which allow only microbes that do not cause disease in humans) that are legitimate and documented, but self-regulated. There are biohacker labs in most major cities of Australia. However, there is also biohacking in the community outside of DYI lab network, including home gene-editing. With citizen science, there may be risks such as illegal disposal of biological waste, or operating at higher biosafety levels than allowed, and unless we have methods for tracking such activity, quantifying and mitigating the risk is difficult. What happens to waste streams in labs is a concern, as dumping into the environment can affect humans, animals or plants, can accumulate in soil, spread through waterways or be aerosolised, depending on the method of dumping. Issues with waste can even occur in the most prestigious official labs, as seen in the incidents at USAMRIID, which resulted in waste water being sprayed into the surrounding environment. So what then of the undocumented labs that may exist in the community?
Illegal labs
Some illegal labs may have nefarious intent and aim to stay under the radar in the community. Just as home meth labs have proliferated in the community, so too is it possible that illegal biolabs exist in the community. A lab in a box kit can be purchased online, as can genetic code and reagents. The open access publication of methods for gain-of-function research and synthetic biology, coupled with the rapidly declining cost of doing such research, poses an increasing risk of lab-generated epidemics or pandemics. The stakes are higher now than even a decade ago because of the accessibility of technology in gain of function and synthetic biology. The scientists that created an extinct orthopox virus (closely related to smallpox), in a lab say “The advance of technology means that no disease-causing organism can forever be eradicated.” Synthetic biology, which includes hundreds of private companies, is also self-regulated under a voluntary code of conduct.
Reedley incident a wake up call
The finding of a clandestine laboratory in Reedley CA, in the US, should be a wakeup call for intelligence and law enforcement agencies. It was discovered in December 2022, purely by accident, when a local code enforcement officer noticed a garden hose coming out of the wall of an unassuming warehouse building (a red flag for identifying drug labs). When authorities searched the building in March 2023, they were shocked to discover a laboratory, nearly 1000 genetically modified mice, unlabelled test tubes and containers full of biological material and chemicals. Pathogens including SARS Co V2, HIV another BSL 3 agents were reportedly found in this illegal lab. Ostensibly the company, Prestige Biotech, was making COVID tests and other medical products. The lab was moved to this site after a predecessor company, Universal Meditech, Inc. went bankrupt. There were already red flags about Universal Meditech, a sister company of a Chinese company, when in late 2022 their SARS Co V2 rapid antigen test was recalled by the FDA because “products were distributed without appropriate premarket clearance or approval which potentially could result in inaccurate test results due to lack of performance evaluation by the FDA.” An even earlier red flag was a fire in the lab in 2020. Prestige Biotech did not have a licence to operate in Reedley nor did they have a contract to dispose of licenced medical waste.
The Reedley incident raises several issues. The first is community safety. Did illegal dumping of waste, or accidental infection of an employee, cause any outbreaks? Second is the ethics of keeping animals in inhumane conditions. Third is the question of what was actually being done in the lab. Fourth is the ability of law enforcement and intelligence agencies to think outside the box and actively detect such operations. Monitoring for sales of high consequence pathogens or biosafety cabinets is not enough. The array of potential pandemic pathogens is vastly greater today than it was during the Cold War or even a decade ago, thanks to quantum advances in technology and the convergence of different and disparate technologies. The next issue is whether regulation and compliance monitoring can be improved, especially once a company is red-flagged – in this case it should have been red flagged after the fire and again after the FDA recall. If it is known a company using biological materials has closed, what is the process of ensuring safe and lawful disposal of lab materials? How was it Universal Meditch, Inc. could simply move their lab, animals and materials to a warehouse? And why did it take 3 months to get a warrant to inspect the premises? Finally, it raises the issue of national security.
Trojan horse
What if a company with ties to a hostile state was found on domestic soil making a nuclear bomb under the pretence of making automobiles? Such a deception would be much easier with biological than nuclear agents. I am not saying this was the case with Reedley, but surely a Trojan Horse approach is a real risk, especially if governments have no systematic method to track illegal labs. A hostile state may establish labs doing some kind of seemingly legitimate work within a target country, but use that as a shopfront for more nefarious work. Prestige Biotech, which did not have a valid Californian address, but has addresses in China, may have been simply making COVID tests, with no ill-intent. However, this case illustrates how easily nefarious actors can set up clandestine biolabs and escape detection.
Playing catch up
In the cybersecurity field, governments are constantly playing catch up with nefarious actors, who have the latest technology earlier and without any regulatory barriers. Biosecurity is about a decade behind cybersecurity in terms of preparedness and response capacity of intelligence and law enforcement agencies, because there is less awareness of the threats posed by biological technology and the increasing accessibility of this technology. It’s time to wake up and change our approach to detecting illegal labs. We do better with illegal drug labs, so we can at least use the same approach to illegal biolabs. The stakes are arguably higher because a drug lab will not cause a pandemic, but a biolab may well.
A starting point would be to turn to law enforcement to develop some immediate solutions around illegal biolabs. They can build on first responder preparedness for illegal drug labs by adding training, PPE and protocols for biolabs to the potential hazardous environments that frontline police or other agencies may stumble upon. Next, law enforcement and intelligence agencies need proactive systems to detect illegal biolabs. Open-source data alone has the potential to warn of epidemics and other hazards. The informatics technology available today can harness vast textual, video and other data streams using AI to generate signals of potential biolabs. Even better if non-open data (such as data on procurements, emergency department visits for possible toxic, chemical or infectious exposures, mobile phone signals, travel and movement data) could be harnessed to detect potential illegal labs. This needs inter-disciplinary strategic thinking, linkage of critical infrastructure systems, harnessing information well beyond microbiological data, use of advanced informatics and AI methods and political will.
We need new approaches to biopreparedness, and there is so much more we can do. The task ahead of us, however, is vast, and requires a truly inter-disciplinary approach involving health, law enforcement, intelligence, defence, emergency response and community. Another essential step is raising the awareness of the community, who is the most important stakeholder in either research to develop drugs and vaccines, or the risks of lab accidents. Our research showed very low community awareness of gain-of-function research and dual-use biological technology. When people were provided facts about such research, the majority found the risks unacceptable. Yet the debate about risky enhanced potential pandemic pathogen research has been tightly held within science and policy circles, with little engagement of the community. Engagement of the community, law enforcement and intelligence is more rapidly achievable than the change required to achieve inter-disciplinary new approaches to biosecurity. These are starting points that can begin to mitigate a very real risk of illegal labs in our communities.
OSINT for surveillance
Meanwhile, early warning systems for serious epidemics are needed.
What is OSINT?
We live in an information age, with enormous amounts of publicly available data on the Internet. We need AI tools to harness this information. Open-source intelligence (OSINT) is intelligence extracted from vast, untapped publicly available information. Since 9/11, law enforcement agencies have used OSINT to capture early warnings of planned terrorist attacks. Public health has been slower to recognize the potential of OSINT. OSINT requires the use of AI to make sense of vast quantities of unstructured data, as simple web scraping will result in a massive data dump containing too much junk to be useful. EPIWATCH® is an Australian innovation that uses three AI systems to make sense of the data and provide curated, valid outbreak data. The public dashboard provides general data, but the internal dashboard has greater capability.
OSINT is valuable for biosurveillance because it can provide an early warning of an emerging epidemic before governments know about them. Formal surveillance from laboratory diagnostics or notifications from doctors is typically delayed. Yet true epidemic diseases grow exponentially, so timely early warnings are critical to our chances of stamping out or preventing spread from an early epidemic. By the time COVID-19 was acknowledged as an epidemic in Wuhan, in January 2020, it had already spread around the world. Serological studies show that SARS Co V2 was circulating in the USA in December 2019 and in Europe in November 2019. If other countries could have identified an OSINT signal for severe pneumonia in Wuhan in November or even December 2019, independent of data released from formal government or other authorities, they would have been better placed to prevent or mitigate spread across their borders. OSINT must be multilingual to capture global epidemic signals. We have shown in our research that the majority of signals will be missed if searching is only in English. In fact, 70% of the intelligence in EPIWATCH® is from non-English sources. EPIWATCH® collects OSINT in 46 languages, including all major languages of Asia and the Pacific.
OSINT provides the earliest warnings because it taps into local communities and local news agencies and provides warnings about epidemics long before health authorities are aware of them. In some cases there may be financial disincentives to testing and reporting (eg. due to impacts on trade, tourism or farming), and OSINT can overcome censorship or reporting delays. A good example is H5N1 avian influenza in dairy farms, where there is a disincentive to farmers testing or reporting. In addition, new pathogens for which there is no test yet can be detected by OSINT. This can be done by searching for clinical syndromes such as pneumonia, encephalitis or rash and fever, which will be the first sign of new infection emerging.
OSINT Is not a replacement for formal surveillance, but an enhancement, which can trigger a much earlier response, investigation and diagnostics. It can also enable targeted surveillance and diagnostics. For example, wastewater surveillance depends on the location of the sites and cannot pick up epidemics occurring in other locations. It also cannot pick up diseases for which no test yet exists. OSINT can enable Geo targeted wastewater surveillance, biosensors or formal diagnostics by providing a location likely to provide high yield for testing. OSINT can also enable earlier characterization of pathogen genomics, which can be invaluable for development of vaccines and diagnostics. Ideally, OSINT like EPIWATCH® should be used as part of an integrated, end to end bioinformatics pipeline for optimal biosurveillance. No country has this as yet, with separate systems (such as wastewater surveillance, lab notifications, sentinel surveillance) operating in parallel without integration of data.
Finally, in the age of synthetic biology, genetic engineering and DIY biohacking, there is a risk of unnatural epidemics. The question “Is this outbreak natural or unnatural” should be asked of all outbreaks in this day and age. If the question is never asked, unnatural outbreaks will never be identified, as I outline in my book, Dark Winter. Historical unnatural epidemics include the Sverdlovsk anthrax leak and the “Russian flu” of 1977. The discovery of an illegal biolab in Reedley, California in 2022 also highlights the risk of unsanctioned, dangerous research that may result in a lab leak. Risk analysis tools can be applied to any epidemic to determine a likelihood of natural or unnatural origin. We have done research on the best available tool, the Grunow Finke tool, to improve sensitivity and specificity. Yet such tools are never or rarely used in public health, and nor is the question of origin of an epidemic asked. We are still a long way from being truly prepared to prevent and mitigate both natural and unnatural epidemics.