Terrorism threatens America's computer systems and public health, but the race to put defenses in place threatens to overrun traditions of open scientific research, leading science policy thinkers warn in a new report from the American Association for the Advancement of Science.

"It all depends on how this homeland security thing plays out," Lewis M Branscomb — former director of the science, technology and public policy program at Harvard University's John F Kennedy School of Government, and one of the contributors to Science and Technology in a Vulnerable World — told The Scientist.

"We could for the first time have a combination of two effects, one being a significant research agenda funded for reasons of national security, which in history has meant generous funding," Branscomb said. "On the other side, the research activities would be almost entirely within the civilian sector, very different from the Cold War situation where the big money went to the military–industrial complex and the university work was merely a by-product."

In the report, Branscomb adds his voice to those of University of California, Santa Cruz Chancellor MRC Greenwood and Massachusetts Institute of Technology Professor of Political Science emeritus Eugene B Skolnikoff, who wrote that research scientists need to be wary of security policies that threaten the core of the scientific enterprise.

Skolnikoff told The Scientist that government moves in the months since the spring meeting "leave me very concerned about how the new department is going to carry out its responsibilities with regard to scientific information."

A great deal of university research may end up hidden away bearing the uncertain label of "sensitive but unclassified" information, Branscomb wrote in his contribution to the report, which was the product of a spring colloquium.

Greenwood pointed out that the fledgling Office of Homeland Security was quick to cast a suspicious eye at the traditions of open publication and access to data sets that are essential to the replication of scientific results. The three authors also cited government surveillance of foreign students and restrictions on overseas communications as potential problem areas.

Although tracking students could offer some protections, Greenwood wrote, "restricting international students would contribute to an international perception that we are isolationistic and imperialistic — and that would not be a good thing for national security either."

"As far as I know, no student who has been admitted legitimately to a research institution in the US has been involved in any terrorist activity," Greenwood said.