The government is developing a scientific code of conduct to help prevent legitimate research being employed by terrorists.
Leading academics have warned scientists that they must agree a system of self-regulation or risk state interference in their work.
Fears have been raised by the publication of a select committee report on the scientific response to terrorism and the aborted attempt by the US government to get an international ban on therapeutic cloning.
Scientists on both sides of the Atlantic are drafting voluntary codes of practice. These could forestall any government bid to constrain science in an effort to prevent research in fields such as infection and immunology being exploited to create a terrorist weapon.
But it emerged this week that the Foreign and Commonwealth Office has already taken a lead in directing the conduct of researchers.
Mark Walport, director of the Wellcome Trust, said: "Scientists have responsibilities. If the scientific community is resistant to self-regulation, it can't complain if governments with legitimate concerns decide to intervene."
A statement issued by the Wellcome Trust backs the self-regulation of the creation and dissemination of scientific knowledge and notes that it has begun developing guidance for reviewers and applicants for its own work.
It adds: "Overzealous government intervention, no matter how well intentioned, could result in the decline of international collaborations and may discourage talented overseas scientists from coming to work in the UK."
The science and technology committee report, published last week, calls on research councils and learned societies to set up a code of conduct for scientists and to enforce it by refusing grants or membership to researchers who do not comply. Noting growing concern in government about the regulation of science, the MPs say the research community risks having "ill-judged restrictions" placed on it by politicians if it is not seen to act.
Nick Winterton, executive director of the Medical Research Council, confirmed that the FCO had assumed responsibility for the code and that preliminary talks between civil servants and the research councils to discuss the issue had been organised. "The research councils will want to play a constructive part in taking this forward with a view to helping to deliver something of practical value," he said.
Ian Gibson, chairman of the committee, warned that the FCO would be unlikely to see the welfare of science in this country as its primary concern. He told The THES : "They need to make sure they don't shoot from the hip and do it on their own. They must consult the scientific community properly."
The idea of a scientific code of conduct has been greeted with scepticism by key science and higher education bodies. Both Research Councils UK and Universities UK told the committee that they were not convinced of the need for this further layer of bureaucracy.
Brian Eyre, chair of the Royal Society's standing committee on the scientific aspects of international security, said a code would have real problems: "At the extreme, the code could start defining the research agenda. How do you assess whether a piece of fundamental research will eventually be a threat?"
A code would be useless if it were not applied internationally, he said, but he advised the UK not to try to emulate the more draconian controls that have been introduced in the US.
A spokeswoman for the Royal Society said: "Any government legislation to limit research should be applied in a careful and sensitive manner to ensure that scientific research to combat terrorism is not hampered."
Representatives of the Royal Society's science advice section hope to meet members of an expert panel of the US National Research Council, which has produced a report calling for self-regulation to prevent the misuse of biological research. This identified seven experiments that would be of particular concern:
* Rendering a vaccine ineffective
* Conferring resistance to therapeutically useful antibiotics or antiviral agents
* Enhancing a pathogen's virulence
* Making a pathogen more contagious
* Enabling a pathogen to infect new species
* Enabling a pathogen to evade detection
* Making a biological agent or toxin useable as a weapon.
Mike Crumpton, a founding fellow of the Academy of Medical Sciences who gave evidence to the committee, said an ethical code would raise awareness of terrorist issues among researchers, but he noted that monitoring adherence would be "virtually impossible".