Even before the fatal anthrax mailings in 2001 were traced to a government lab worker, US regulators had been grappling with how to change to the rules governing research with dangerous pathogens so as to keep hazardous agents out of the hands of would-be bioterrorists. One solution, suggested last month by an intergovernmental working group, is for biological agents considered to have high potential as weapons to be stratified on the basis of risk.

The group's report, released on 7 January, recommended dividing the current list of 'select agents'—which contains 82 pathogens and toxins—into several tiers of security. The working group found that such a tier system would further research on agents judged to be less of a threat by requiring less bureaucracy and fewer costly security measures. Simultaneously, it would encourage research institutions to focus more resources on agents with the greatest potential for harm.

“What everyone seems to recognize is that not all of these agents are created equal,” says working group co-chair Carol Linden, the principal deputy director of the US Biomedical Advanced Research and Development Authority. “Right now, there is a somewhat blanket approach to security for everything. That isn't necessarily providing the best security for the most dangerous agents; meanwhile, it's needlessly preventing important work on others.”

The panel does not recommend a specific grading system; instead, it calls for one to be created through a collaborative effort between the government agencies that currently regulate select agents and public agencies such as the National Academy of Sciences (NAS) and the American Society for Microbiology.

However, legislation that is currently awaiting a vote in the Senate has proposed a three-tiered security system, with the most hazardous agents regulated as potential weapons of mass destruction by the Department of Homeland Security. Although the legislative move to refine the list is widely applauded, this highest level of security has caused concern among many scientific organizations.

“It's a mistake to treat these agents like nuclear weapons,” says Gerald Epstein, a security policy expert at the American Association for the Advancement of Science (AAAS) in Washington, DC. Rather than adding additional safety gauges, he says, the Senate's proposed oversight system would mainly impede research by adding unnecessary costs and red tape.

In November, the AAAS sent an open letter to the senators who sponsored the bill, calling on the government to work within its existing framework while attempting to foster a network and culture of self-regulation similar to the US National Institutes of Health's biological surety program, which covers all personnel who work in biosafety level 4 laboratories.

“What is important is to move the burden of responsibility away from a static list and toward real-time evaluation and judgment,” says David Relman, an infectious disease researcher at Stanford University School of Medicine in California who chairs the NAS forum on microbial threats. Last month, Relman coauthored a perspective article arguing that the select agents list, in its current incarnation, hinders research into vaccines and therapeutics, thereby making society more vulnerable to biological attacks and natural epidemics (Nat. Rev. Microbiol. 8, 149–154, 2010).

A list of dangerous agents is inherently flawed, Relman says, in part because pathogenic strains vary greatly, even within a single species. What's more, they can be genetically altered to either be dangerous or benign. Instead of a rigid classification scheme, systematic ways of evaluating the potential danger of organisms on a case-by-case basis need to be developed, he adds.

“The truth is that we're doing our best—a more intelligent, stratified list is a step in the right direction,” he says. “We don't have the technology or the knowledge right now to evaluate everything we work with, but that's a goal worth working toward.”