On Thursday, biosecurity researchers released the first comprehensive map of the world’s maximum biological containment labs. Those facilities, known in the US as Biosafety Level 4 (BSL-4) labs, work with the deadliest pathogens under high-security precautions, from pressurized “space-suits” to air-locked labs, and until now, had never been published in a central list.

In a Thursday webinar, the authors outlined a series of recommendations for using the map to develop a global approach to biosecurity.

“The aim of our project is to increase public knowledge about Biosafety Level 4 labs, and importantly, to strengthen national and international virus management policies,” Filippa Lentzos, a biosecurity researcher at King’s College London, said on the call.

That means everything from up-to-date education on how to deactivate a live virus, to better protocols for reporting accidents when they do occur.

“Clinical work on pathogens are important for public health, biomedical advances, and disease prevention,” Lentzos continued. “But some of these activities pose significant risks,” both to lab workers and to the public. “There are also security risks that pathogens could be stolen from a lab, or a lab insider could use their knowledge, skills, and access for malevolent purposes.”

To be clear, in spite of this week’s media froth, there’s no biological evidence that ties SARS-CoV-2 to such a lab. Studies of the virus’s genome suggest, though indirectly, that it was neither bioengineered nor cultured by humans. Though many virologists say it’s possible a previously undetected SARS-CoV-2 could have infected a researcher, there’s widespread agreement that a spillover unrelated to a lab is more likely. Coronaviruses spread from animals to humans with frightening regularity—a new one, from dogs, was discovered in Malaysia just this month—while there is no record of a novel pathogen coming out of a lab. And sorting out the precise origin story of the 2001 coronavirus outbreak, SARS-1, took years, even under much less fraught geopolitical circumstances. (It involves wild bats and caged civet cats.)

Still, that doesn’t mean that lab safety isn’t a pressing issue. Regardless of where SARS-CoV-2 emerged, the risk of lab accidents has risen with the number of labs. According to the report, 21 maximum biosecurity labs have been established in the last ten years. The actual number might be higher—the researchers couldn’t find founding dates for about 20 labs. According to the report, three quarters of these are in urban areas, raising the risk that a mistake could turn into an epidemic.

And high-risk incidents have happened at these labs. For the most part, they involve a BSL-4 lab shipping samples of a pathogen, like Ebola and Marburg, that had not been fully inactivated to a less secure facility. In 2004, one person in Beijing died of SARS-1 after her daughter, a virologist, was infected in a lab accident. (This instance didn’t take place in a BSL-4 lab, and SARS isn’t categorized as a highest-risk pathogen.)

[Related: The fight to stop the next pandemic starts in the jungles of Borneo]

The problem with trying to count these facilities is that there’s no clear international definition of a BSL-4 lab. On the borderline are labs that might have the capacity to scale up to BSL-4 in a crisis, or mobile BSL-4 labs that could perform autopsies in an anthrax outbreak. There are also no binding agreements for safety precautions in the highest-security labs, though there are a series of recommendations from the World Health Organization.

The map released Tuesday includes 59 labs, along with preliminary information on their biosecurity protocols. The majority are in Europe, although they’re spread across the European Union, Russia, and various non-EU countries. There are 14 in North America, and 13 in Asia.

Of the 23 countries that have BSL-4 labs, only 17 have national biosafety associations or are members of international partnerships. International bodies, the researchers say, should be providing support for the remaining countries to develop their own legislation and internal oversight in line with global standards. “It’s not enough for a single lab to have its own policies,” says Gregory Koblentz, a coauthor and biodefense researcher at George Mason University.

But that means some organization needs an international mandate to oversee biosafety. On the webinar, Lentzos raised two possibilities: the World Health Organization, or the UN’s Biological Weapons Convention. “I think neither institution, as it is currently configured, is set up to do this kind of work,” she said. “A more natural fit would be with the WHO, but the WHO doesn’t have this kind of mandate. So some kind of restructuring would need to happen.”

That kind of international collaboration could be moving further from reach as American politicians seek to pin blame on China for COVID-19, making it harder to prevent a future lab-related outbreak by increasing tension about the issue.

The other significant challenge that would face such an international body is developing a clear risk assessment of different research types. Lentzos referred to the recent focus on “gain of function research,” in which microbes are altered in a laboratory to study how they might become human pathogens. “From our perspective, gain of function is just one type of potentially high-risk research. Just looking at whether or not a lab is carrying out gain of function research doesn’t tell you so much. … You need to identify high-risk research more generally.”

And that means identifying other, lower-security labs that might also be doing high-risk research. This project, the researchers say, focused on a very narrow category of labs only as a starting point. They’d like to extend the work to cover BSL-4 labs that deal exclusively with animal pathogens, as well as into BSL-3 facilities, which work with slightly less deadly, but extremely dangerous, pathogens like tuberculosis and plague.

“When we talked to colleagues earlier about the study, they would say, okay, you’re mapping the BSL-4 labs. But is that really where your risks are?” explained Lentzos. “We completely agree. BSL-4 labs are the very tip of the iceberg.”