In suggesting that biologists must regulate their work to prevent the free flow of information to would-be bioweapons developers, George Poste will win himself few friends at the lab bench. Yet biologists would do well to ponder on the comments of this adviser to the US Department of Defense, and former senior executive in the drugs industry, made at a conference in London (see page 237).

Poste accuses biologists of naivety in failing to consider the possibility that their research findings could be used for malign purposes. Nature expressed similar sentiments in an editorial published in May this year (411, 223; 2001).

But the difficulty is how to respond to the fact that data generated in legitimate projects could have applications in bioweapons development. The anguished reactions to some of Poste's suggestions — such as the vetting of manuscripts to determine whether 'sensitive' findings should be withheld from the open literature — show that there are no simple answers. Although such ideas might not shock cryptographers, who are used to dealing with issues of national security, this is uncharted territory for most biologists.

Scientific societies and research agencies should now rise to the challenge by holding meetings to debate the issues. Poste points to the example of the 1975 Asilomar Conference on Recombinant DNA Molecules, which considered the risks of recombinant DNA research at a time when US biologists were observing a voluntary research moratorium. Out of this came guidelines that were adopted, in modified form, by the National Institutes of Health.

The potential applications of advances in biology to the development of 'enhanced' bioweapons poses more complex problems, however, and it is unlikely that a clear consensus will emerge. It is also important to understand the nature of the threats involved. The sorts of projects that Poste is most worried about — such as the creation of 'stealth' viruses that would evade the immune system — are most probably beyond the technological capabilities of whoever is behind the current mailed anthrax attacks in the United States. It would be inappropriate, therefore, to impose stringent restrictions on such work as a reaction to the immediate terrorist threat.

Such technologies are within the reach of a well-funded state bioweapons programme, however. So, in the longer term, Poste is right to be concerned. The most useful result of the debate he wishes to stimulate would be a heightened awareness among biologists of the potential dangers, which hopefully would influence decisions on whether or not to pursue particularly risky projects in the first place.

When experimenting with biotechnologies that try to simulate and accelerate the natural evolution of traits, for instance, is it really wise to test them by heightening the efficiency of genes that confer bacterial resistance to antibiotics, as some researchers have done? The same question can be asked of work that splices genes for cytokines —proteins that manipulate immune responses — into potentially pathogenic viruses.

Some self-restraint is desirable when pursuing the frontiers of biology. Discuss.