Why don't people resent driving licences? After all, most drivers don't intend to go out and kill people and they don't like their freedom to roam restricted. But road vehicles can be turned into a means of destruction, whether through carelessness, mischief or malice. People accept, therefore, a licence that reinforces rules designed to help keep people safe.

Substitute ‘organisms’ for ‘road vehicles’ and ‘researchers’ for ‘drivers’ and you have a case for the licensing of biologists. And however outrageous Nature readers may consider it, politicians and policy-makers are taking codes of conduct and licensing in research seriously. This much was clear from last week's inaugural meeting of the US National Science Advisory Board for Biosecurity (NSABB). Just a US problem? Applicable only to defence labs? Think again. Any biologists whose work could be misused (and they are many) face growing pressure to reassure their fellow citizens. This would also pre-empt excessive regulation by a willingness to declare explicitly just how responsible they intend to be.

Most researchers would wonder what planet such proposals come from. Speaking at the NSABB meeting, a social scientist, Brian Rappert of Exeter University, UK, described discussions with 600 biologists in the United Kingdom, most of whom were blissfully unaware of issues of ‘dual use’ that learned societies, biological weapons convention negotiators and others have been fretting about for years. Speakers from the intelligence services and weapons inspectorates emphasized, however, the active interest in the harmful uses of biology they have encountered among a malevolent few.

The approach favoured at the NSABB's preliminary discussions was to all but accept that genies cannot be kept in bottles, and to focus instead on developing a ‘culture of responsibility’ in research institutions, journals and, especially, among individuals (see Nature 435, 860; 2005). Engineers and medics are imbued with a culture that, if ignored, can lead to professional prohibition. For most basic researchers, codes of behaviour, although taken for granted where safety and human or animal experimentation are concerned, are alien when it comes to broader professional practice.

If codes are to carry any weight they will need to be backed by certification, and institutions will have to extend their means of assuring compliance.

Consider the phrase ‘do no harm’. Deceptively simple, a trite piece of motherhood and apple pie, and yet, as one medical researcher at the meeting said, this fundamental principle had provided him with significant help when faced with some critical professional decisions. As Rappert highlighted, many codes for biologists have been drafted, ranging from such statements of aspiration to enforceable codes of practice (see http://www.projects.ex.ac.uk/codesofconduct).

If they are to carry any weight, they will need to be backed up by certification, and research institutions will need to extend their often flimsy means of assuring compliance. Less demandingly, good behaviour can be encouraged by education, as illustrated at the meeting by Nobel laureate Phillip Sharp with examples of courses on science ethics and best practice at the Massachusetts Institute of Technology.

Codes of practice have so far attracted little attention in the biology community. But in a world threatened by terrorism, governments are taking more interest in such codes, and scientists would do well to engage in a constructive discussion about what role they might play.