The US Centers for Disease Control and Prevention (CDC) has published an interim rule placing the reconstructed 1918 flu virus on its list of select agents, and outlining provisions for its safe handling. But these are just the first steps that need to be taken to assure the public that the virus is in safe hands.

The interim rule, which was published in the Federal Register on 20 October, means that the virus may be shared with laboratories in the United States that have registered with the agency (see Deadly flu virus can be sent through the mail). Some sharing is needed to accelerate progress in understanding its virulence — but it will also increase the risks of an accidental release. The classification of the virus is welcome, although some virologists would argue that it is overdue, given that the existence of the strain was well known months in advance of its publication (T. M. Tumpey et al. Science 310, 77–80; 2005).

The 1918 flu virus is hard to contain and is capable of spreading rapidly between people. The researchers who work with the reconstructed virus point out that current flu vaccines and drugs provide good protection from it — but these are in short supply, and the threat of an accidental release is real.

The risks of such release during the physical shipping of the virus will be reduced if laboratories choose to construct it themselves, on the basis of the published sequence. But that still leaves the risk of an escape from labs that work with it.

The CDC has ruled that enhanced biosafety level 3 laboratories can work with the virus. That decision seems justifiable, in the interests of rapid research.

The CDC has ruled that enhanced biosafety level 3 laboratories can work with the virus, rejecting calls for a tougher, level-4 requirement that would have restricted the work to a handful of laboratories. That decision seems justifiable, in the interests of rapid research.

But uncertainty continues to cloud the question of access to the virus for laboratories abroad, where the CDC's writ doesn't run. Already, a biosafety level 4 lab in Winnipeg, Canada, has announced plans to reconstruct the virus.

No one will question the motives or the security arrangements at the Canadian lab, but the question of international regulation for this and other reconstructed viruses remains fraught. There is no international regime for the mandatory regulation of virus reconstruction, and it is hard to imagine how one could be put together in the time available.

In 1994, however, the World Health Organization (WHO) brokered an agreement restricting the smallpox virus to just two laboratories across the world. National governments should ask the WHO to examine the need for a broader agreement between member states to oversee the distribution of potentially dangerous, reconstructed viruses such as 1918 flu.