51¹ú²úÊÓƵ researchers are being frustrated by restrictions prompted by the US's war on terrorism that set limits on how they conduct their work - and who is allowed access to it.

One of the issues is the addition to the category of "classified" of a new category called "unclassified but sensitive", which applies to certain research. Another new law gives the government control over whether foreign scientists can be involved in government-funded research.

Scientists themselves are also imposing restrictions- 32 scientific journals announced that they would not publish information that might help terrorists.

These measures have attracted scorn from some academics, though only a few have spoken out. "I believe that restriction is rarely advisable, and certainly rarely feasible, in this environment," said Charles Vest, the president of the Massachusetts Institute of Technology, which turned down a $400,000 (?256,000) computer science grant from the National Security Agency rather than provide information to the government about foreign researchers working on the unclassified project.

51¹ú²úÊÓƵ

"Restrictions on our teaching and where our students come from are unlikely to counter (national security) concerns," Mr Vest said.

Such opposition has not slowed government involvement in academic research.

51¹ú²úÊÓƵ

A Czech-born 51¹ú²úÊÓƵ of Connecticut graduate student, for example, was charged under the USA Patriot Act with illegal possession of a select biological agent for placing two vials of anthrax in a locked laboratory freezer for later research after his supervisor told him to destroy them.

The laws also regulate unclassified but "sensitive" information, a term that first appeared in a memo by president George W. Bush's chief of staff and was then added to the 51¹ú²úÊÓƵland Security Act, and which has mainly created confusion about what university research can and cannot be disclosed.

"The problem is that nobody knows what that is," Ms Polmar told a national science radio programme. "We know what classified is and we know what unclassified is, but there are no standards or definition for what sensitive but unclassified is. We don't know how to deal with this one."

Some universities have vowed to turn down government money rather than subject themselves to the new restriction on the publication of research.

51¹ú²úÊÓƵ

Cornell 51¹ú²úÊÓƵ, for example, which received more than $360 million last year from the federal government, said it would not accept any contracts that required "prior restraint" of information.

Cornell has already turned down a $400,000 grant because the government demanded that it be allowed to review the findings before publication. The topic: sexual assault and harassment of international students in the US.