Complementing Biosafety Levels With Societal Safety Levels
The paper proposes complementing existing Biosafety Levels with new Societal Safety Levels (SSLs) that quantify expected public fatalities to ensure high-risk pandemic pathogen research meets enforceable safety thresholds comparable to other hazardous industries, thereby better protecting the non-consenting public from potential laboratory releases.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer
Every time a scientist works with a dangerous germ in a laboratory, they follow a set of rules designed to keep that germ inside the building. These rules, known as biosafety levels, act like a series of increasingly tight locks on the doors. The system was built with a specific goal: to protect the people working inside the lab from getting sick. For decades, this approach has worked well. The scientists who handle these microbes are trained professionals who understand the risks, and the safety measures have successfully prevented them from becoming victims of their own work.
However, a critical question has gone largely unanswered: what happens if a germ escapes the lab and reaches the general public? The current rules focus almost entirely on the worker, not the neighbor. They do not fully account for the fact that some germs can spread from person to person, potentially causing an outbreak that sweeps through a city or even the world. A virus that kills a single person in a lab might be contained easily, but a virus that can jump from one person to another could cause a pandemic. The existing system treats these two very different dangers with the same level of caution, which means that some of the most dangerous threats to the public might not be getting the strictest protection.
A team of researchers at the Massachusetts Institute of Technology has proposed a new way to think about these risks. They argue that we need a second set of safety ratings, one that measures the potential harm to society rather than just the risk to the scientist. They call this new system Societal Safety Levels. Instead of just asking, "Is the scientist safe?", this new framework asks, "If this germ gets out, how many people could die?" By using numbers to estimate the total number of expected deaths from a potential accident, the researchers hope to create a standard that matches the safety rules used in other high-risk industries, like aviation or nuclear power.
The researchers began by looking at the history of laboratory accidents and natural outbreaks. They found a strange mismatch in how we currently classify danger. For example, the 1918 influenza virus, which killed between 50 and 100 million people and could cause a massive pandemic today, requires fewer safety precautions in a lab than many non-infectious agents that cannot spread beyond the person who is directly exposed. A non-infectious agent might kill a scientist if they breathe it in, but it cannot jump to anyone else. In contrast, the 1918 flu could infect billions. Yet, under current rules, the non-infectious agent often requires a higher level of containment than the virus that could end the world.
To fix this, the team developed a method to calculate the "expected harm" of any research project. They looked at three main things: how likely it is that a germ will accidentally escape the lab, how likely it is that the germ will spread widely if it does escape, and how many people would die if a large outbreak occurred. They combined these factors to estimate the number of deaths a single laboratory might cause in a year. They then compared this number to the safety standards used by other major industries. In fields like dam safety and chemical manufacturing, regulators have agreed that a facility should not be allowed to create a risk of more than 0.1 expected deaths per year. This means that if a factory or a dam is so dangerous that it is expected to kill one person every ten years on average, it is considered too risky to operate without major changes or special permission.
When the researchers applied this same 0.1 deaths per year standard to biological research, the results were startling. They found that many studies involving potential pandemic pathogens, such as historical strains of influenza or newly discovered bat coronaviruses, would far exceed this limit. For instance, research on the 1918 flu virus, if it were to escape, could result in thousands of expected deaths per year, making it vastly more dangerous than the safety limits allowed for a nuclear power plant or a large chemical factory. Even studies on viruses that are currently circulating, like SARS-CoV-2, would need to be re-evaluated. However, the researchers noted that once a virus is already spreading widely in the human population, the risk of a lab accident adding to the total number of cases becomes negligible, because the virus is already everywhere.
The paper proposes a new five-tier system to manage these risks. The lowest levels would apply to agents that pose almost no risk of spreading, requiring no extra rules beyond what is already in place. The middle levels would require more oversight, such as mandatory reporting of accidents and reviews of the benefits of the research. The highest level, reserved for the most dangerous experiments, would require specific authorization from an international body, similar to the strict rules that currently govern research on smallpox. This top tier would apply to experiments that try to make a virus more infectious or studies on historical pandemic strains that could easily start a new outbreak.
Crucially, the researchers emphasize that this new system does not mean stopping all dangerous research. Instead, it encourages scientists to find safer ways to do their work. They point out that many medical breakthroughs can be achieved without using live, infectious viruses. Scientists can study the genetic code of a virus, use dead versions of the virus, or use computer models to design vaccines and drugs. These methods can provide the same scientific answers without the risk of a pandemic. The new framework would push researchers to use these safer methods whenever possible. If a scientist really needs to work with a live, dangerous virus, they would have to prove that the benefits are so great that they outweigh the massive risk to the public, and they would have to show that no safer alternative exists.
The authors also suggest that this approach could actually improve safety for the scientists themselves. By focusing on the risk of spread, the system might reveal that some agents currently locked in the highest level of containment are actually safe to study with fewer restrictions, as long as they cannot spread. This would free up resources to focus on the truly dangerous agents. The goal is to bring the life sciences in line with the rest of the industrial world, where safety is not just about protecting the worker, but about ensuring that the public is not exposed to unacceptable risks. By setting clear, numerical limits on how much harm a laboratory can cause, the researchers hope to rebuild public trust and ensure that the pursuit of scientific knowledge does not come at the cost of global safety.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.