Pentagon advisors urge caution in tightening science security

Despite rising worries about sensitive US technologies leaking to China, a special government advisory group urged caution before imposing any broad new controls over fundamental science, lest they inhibit important discoveries.

The National Science Foundation “should proceed with caution before adding access or dissemination controls to grants or contracts,” the group, called JASON, reported on Thursday. It rejected the idea of designating specific scientific sub-disciplines as risky. That’s the kind of broad-brush security approach adopted by the Canadian government in January.

The JASON group, formed in 1960 as the Cold War intensified, brings together several top American scientists in a confidential advisory body to the Pentagon and security services.

Since at least the Reagan administration, the US government position has been that fundamental research should be restricted only in special cases of national security classification – and the JASON group, noting that now “is not the first time that a national debate has been raised on the issue of research security,” said that those 1980s-era “principles remain valid” today.

Rather than a broad rule change, the group in its unclassified report urged a project-by-project consideration of any security risks at the time scientists apply for an NSF grant. If a proposed research project appears “sensitive”, it said, there is a wide range of steps short of restricting access that the NSF could impose to protect the results, such as revising the proposal, demanding more-frequent progress reports, or requiring special cyber-security safeguards. In doing so, the agency “should weigh the balance” between security controls and open science, the group said.

It specifically rejected the idea, advocated by some in the US security establishment, of writing broad new rules for “controlled unclassified information” – a term for research that, while not legally classified as secret, is still subject to export or other forms of control. It called such controls “a rather blunt and ineffective tool” for security, that “should not be used as a one-size-fits-all approach to mitigating research risk.”

Since at least 2019, US concerns have been rising over militarily or commercially sensitive research results leaking to unfriendly countries, especially China. In 2022, Congress ordered tougher measures to protect sensitive research results, and the NSF is currently setting up a new clearing-house of information and guidelines about research security.

Similar security reviews are underway across Western-allied countries – including in the EU, Britain, Australia, Canada, Japan and Korea. In January, the European Commission proposed a series of new measures to improve research security, in large part due to concerns about China.

The JASON report, which was unclassified, was required under new security provisions mandated by Congress in the 2022 CHIPS & Science Act. In a statement, the NSF said it is analysing “the findings and will consider the implementation of the recommendations as it continues to develop and implement new policy review processes for national security concerns.”  Under the law, it has a 22 May deadline for finishing its security review.