UC Berkeley News


Research under fire: In the war on terror, academic freedom could wind up as collateral damage

| 27 January 2005

The University of California's credo, "Fiat lux" ("Let there be light"), celebrates the power of knowledge. Federal officials, however, mindful that power is a double-edged sword, seem intent on imposing an alternate, post-9/11 credo on those who conduct university research: Let there be licenses.

In an age when data can be dangerous, the Bush administration is clearly concerned with keeping classified information out of the hands of America's enemies, from terrorist networks to hostile regimes.

But university officials warn of a dimly grasped threat from the U.S. government itself, having less to do with legitimate security worries than with a needless clampdown on academic freedom — including moves to bar non-citizens, and even foreign-born U.S. citizens, from participating in an ever-expanding list of science and engineering research projects.

"I believe they're told by their superiors, 'You get those [restrictions] in there, and if you can't, let's just not give those people money - they're not patriotic.'"

-Joyce Freedman

(Peg Skorpinski photos)

"We front-line contract negotiators are beginning to see more and more clauses in our agreements that say, ‘You've got to give us the names and addresses and birthplaces of all foreign nationals,'" says Joyce Freedman, assistant vice chancellor for research administration and compliance, of funding deals tendered by federal agencies. "And we fight those very hard. But we see them more and more."

Freedman's staff reviews all grants and contracts proffered to Berkeley scholars — both from non-federal sources and from agencies including NASA, the National Science Foundation, and the Pentagon — and is able to spot troublesome clauses before papers are signed. But that doesn't keep funders from calling on researchers to take "security measures" once a project has begun. Patrick Schlesinger, director of research compliance in the UC Office of the President (UCOP), characterizes these efforts as "trying to get through the back door what they weren't able to get at the front end of the process."

Such informal appeals can seem innocent enough. Yet going along with even the most benign-sounding request for caution can subject an unwitting researcher to onerous licensing requirements under federal export laws, from limits on publication to bans on foreign-born graduate students in the lab.

Dodging the 'deemed export' bullet:
A researcher's checklist

Freedman attributes the government's stepped-up muscle-flexing to "xenophobia," and fears it will "change the face of research" if it continues. Barbara Yoder, contract and grant officer at UCOP, agrees, attributing the crackdown to "paranoia in D.C." that universities are "loose cannons." She calls it "a straight-up direct attack" on researchers' academic freedom.

"This gets to a fundamental principle," says Yoder, asserting the university's autonomy in the broad pursuit of knowledge. "If we're not careful, we will radically alter the nature of the [research] enterprise."

The safe and the deemed

From the perspective of the Bush administration, it seems, some of those responsible for keeping America safe haven't been nearly careful enough.

Under long-established federal law, the State and Commerce departments maintain lists of "export controlled" commodities — those that, for reasons of national security or trade protection, may not be shipped abroad without a license. In general, the State Department has responsibility for defense-related articles and services — everything from communications satellites to explosive devices — while Commerce controls so-called "dual use" commodities: civilian items and technologies with potential military applications, such as encryption software or underwater-vision systems.

But not all exports are physical commodities, nor must they physically leave the country. Many are so-called "deemed exports," or transfers of controlled information and technical data to non-citizens on U.S. soil. This information-sharing is regulated as though an export-controlled commodity were being sent to the foreign national's country of residency or citizenship, and thus requires the "deemed" exporter — whether a university, a private contractor, or an independent research institution — to obtain a license.

There is, however, a crucial exception to the licensing requirement for deemed exports. If the information is in the public domain — including, in the government's phrase, "fundamental research in science and engineering at accredited institutions of higher learning," conducted openly and with no restrictions on publication or dissemination of a project's findings — there is no need for an export license.

UC Berkeley, by refusing to accept limits on the dissemination of information, operates almost entirely under this "fundamental research" exemption, as do many other universities and research institutions throughout the U.S.

And therein lies the problem, as Washington sees it.

In September 2002, a report by the U.S. General Accounting Office was sharply critical of the Commerce Department's system for regulating deemed exports, concluding it "does not provide adequate assurance that U.S. national security interests are properly protected."

A year later, the inspectors general (IGs) of the Commerce, State, Energy, and Defense departments launched a fact-finding mission to major U.S. research institutions, including Berkeley. According to a report by the Pentagon's IG, the audits found that "at least two contractors and one university granted foreign nationals access to unclassified export-controlled technology without proper authorization." Such access, the report continued, "could allow foreign nations to counter or reproduce the technology," which in turn could "degrade combat effectiveness."

In Freedman's view, the inspectors' findings triggered "[a] kind of intense paranoia" in official Washington about "leaking our secrets to foreign nationals."

Paranoid or not, when the IG's office at Commerce released its final report on deemed exports in March 2004, the title was unambiguous: "Deemed Export Controls May Not Stop the Transfer of Sensitive Technology to Foreign Nationals in the U.S." At its heart were proposals to simultaneously narrow the definition of fundamental research and broaden the definition of deemed exports.

A wake-up call for researchers

 Research files
The campus's files hold research awards that totaled nearly $600 million – about 70 percent of it from federal agencies – in 2004 alone.

The report set off alarms at many of the nation's leading universities and research institutions. The Bay Area Science and Innovation Consortium — a group dedicated to promoting the region's scientific and technical leadership on behalf of major research universities, independent institutions, national laboratories, and businesses — expressed particular concern over a recommendation that licenses be required "before a university could allow a foreign national faculty member, staff member, or student [to] use export-controlled but common scientific equipment like a GPS system, a high-end computer, or a fermenter."

A joint response by the Council on Governmental Relations and the Association of American Universities echoed those concerns. "If Commerce acts as indicated," they wrote, "many fundamental research projects at universities will require determinations of the need for deemed export licenses in order for foreign students, faculty, visitors, technicians, and other research staff to work on such projects. … It will have a chilling effect on university research and education as well as compel discriminatory treatment of foreign nationals on campus."

In a recent letter to Condoleezza Rice, then President Bush's national security adviser (and now secretary of state), MIT President Charles M. Vest — writing on behalf of a group of university presidents and chancellors including UC President Robert Dynes — noted that "significant differences of perception, fact, and interpretation [of deemed-export regulations] persist both in government agencies and at many of the nation's university campuses." The letter adds that "[a] collaborative exploration of these complex matters offers hope to resolve them to the benefit of our national security as well as our leadership in science and technology."

Basing his judgment on a recent meeting between Commerce Department representatives and staff from Berkeley and Lawrence Berkeley National Lab, UCOP's Schlesinger says the agency "seems to have backed off" a proposal to require export licenses based on project participants' country of origin, regardless of their citizenship or residency status.

As for the recommendation to place commonly used research equipment off-limits to foreign nationals, Schlesinger adds, "We're not sure how they're going to come out on that."

Meanwhile, back in the trenches…

Whatever happens at the top of the government food chain, though, mid-level managers from an assortment of federal agencies — agencies that funnel hundreds of millions of research dollars annually to Berkeley alone — are already putting researchers in harm's way.

"The skirmishes are more in the trenches, and the little guys aren't listening to anybody," explains UCOP's Yoder, referring to the program and project managers with whom researchers deal directly.

"I don't believe they really understand it," adds Freedman. "I believe they're told by their superiors, ‘You get those [restrictions] in there, and if you can't, let's just not give those people money — they're not patriotic.' "

And while her office routinely vetoes clauses calling for limits on publication or personnel, even restriction-free contracts can be fraught with peril for researchers. Freedman estimates that she and her staff spend 30 percent of their time trying to negotiate with contract agents to drop restrictive terms and conditions, and working with faculty to help them navigate the increasingly dangerous minefield of export-control regulations.

"If you were a faculty member and your program person called you and said, ‘I think we're in a little trouble, you've got all this stuff up there that might be getting into the hands of the wrong people, take your website down' – of course you would," Freedman says. "The people you're dealing with, you assume they know what they're doing. …Well, the worst thing to do is to take that website down."

Safety first: A checklist for researchers

Officials at UC and beyond say there are several key steps researchers should take to keep their projects free of deemed-export controls:

• Publish early and often to ensure that your research qualifies as "publicly available."

• Do not accept restrictions on access to or dissemination of information, including informal "handshake" agreements with project managers.

• Do not provide citizenship, nationality, or visa status information to project sponsors or other third parties, or agree to background checks for project participants.

• Do not attend meetings from which foreign nationals are barred.

• When in doubt, consult with the campus Office of Research Administration and Compliance (642-0120).

The reason: So long as the project stays in the public domain — that is, the research is conducted openly, and findings are shared regularly with the interested public — the "fundamental research" exemption applies. But the moment a researcher agrees to checks on the publication or dissemination of technical data — or even consents to exclude non-citizens from a "sensitive" portion of a meeting with funders' agents, as nearly occurred at UC Santa Barbara before Yoder intervened — the project is out of "safe harbor," and the researcher may have to apply for an export license. Safe harbor can also be jeopardized by acceding to requests from a government agency, or for that matter a commercial backer, to review materials prior to publication.

‘We're about open exchange…'

In the fall of 2003, Edward Lee, a professor in Berkeley's Department of Electrical Engineering and Computer Sciences, was the principal investigator on a project funded by the Pentagon's Defense Advanced Research Projects Agency, better known as DARPA. The project, which involved generic embedded-software technology, had been designated four years earlier as open research — that is, exempt from licensing or publishing limits — and Lee and his team had routinely posted the results of their work on the department's website.

Yet just as they were about to wrap up the project, the contracting agent asked Lee to add a cover sheet to the final report warning that all information was restricted under State Department regulations — despite the fact that it was already available to anyone with Internet access.

"It seems very odd, after four years, to have them suddenly say that now the information should be restricted," Lee observes. "It makes no sense to do this in the final months of a project."

Lee insists he would never have agreed to do the project without full academic freedom. "Many of our students are foreign-born," he notes, adding: "We're about an open environment and open exchange with colleagues worldwide."

Edward lee

'We're about an open environment and open exchange with colleagues worldwide.'

-Edward Lee

Nevertheless, under pressure from his benefactors, he says he "almost capitulated" to the demand. In the end, Lee consented to a stipulation that the final report itself could not be published — a compromise with no practical impact, since its contents had been posted separately. Still, he says, the agent's anxiety reflects a change in thinking at DARPA, which has cut back significantly on funding for openly conducted research in favor of classified projects.

DARPA, explains Lee's colleague David Culler, "is a very strange place these days." Just months after Lee's brush with export controls, Culler, also an EECS professor, had a similar experience. In 2000 he was awarded an agency contract to develop hardware and software for miniaturized wireless computer networks, utilizing open-source software that would be shared with the wider research community. "This whole notion of openness was fundamental," Culler says. "That's what we wanted to do."

In February 2004, however, DARPA's program manager sent an e-mail to Culler and more than a dozen other researchers working on various aspects of the program, asking that source codes and possibly other material — the message, says Culler, was "ambiguous" — be removed from websites. Unsure of what to do, Culler consulted with Freedman's office, which advised him to take no such action.

DARPA, meanwhile, pondered its next steps, eventually opting to split the program into two major segments, with basic research remaining at universities and classified work going to military contractors like Northrup Grumman. The decision, Culler says, wound up costing Berkeley "very little," though other universities lost "quite a lot."

"The money basically moved from the universities to the military contractors," he says. "It's a tremendous shift in where the resources have gone."

More is at stake than just money, however. Ironically, DARPA's efforts to "short-circuit" the research process — to short-shrift basic research in favor of specific military applications — could have the effect of hampering, not improving, America's security.

"If you're not able to keep the basic-research engine alive," Culler explains, the result is likely to be less innovation and competitiveness, as other countries pick up the slack. "There is absolutely a need for basic research," he insists. "Ultimately, in the long term, that contributes to an advantage in national security."

For the short term, at least, Culler wants his fellow researchers to be alert to the dangers of losing the fundamental-research exemption. "Understanding the implications of losing that waiver is really important for faculty," he says, noting the need to apply for a license and abide by its conditions: "If you violate it [even] by giving a talk to people of other nationalities, you've committed a crime."

Penalties for violating export licenses include fines (up to $1 million for the institution and $250,000 for individuals) and up to 10 years in prison for wayward researchers. Yoder and Freedman, however, are intent on training faculty to resist any and all curbs on publishing and personnel, and thereby avoid the need either to acquire a license or — the more likely outcome, given the university's firm policy against ceding control over research to outside entities — to pull the plug on their project.

"The scare right now is mainly in the engineering community and the computer sciences," says Freedman, adding that she anticipates a further diversion of federal dollars from basic research in these and other dual-use fields into classified and "sensitive but unclassified" projects, which give the government a stronger role in managing research — and researchers.

Meanwhile, says Yoder, export-control rules are "like a background noise that [researchers] have to understand."

Freedom or flight

Already, UC campuses "have turned down millions of dollars in government contracts," reports Schlesinger, noting that other universities are happy to take up the slack. Because "the government needs UC-, MIT-, Harvard-level research," however, he believes the Bush administration will eventually begin to draw a finer distinction between fundamental research and that which truly needs to be classified.

That notion is echoed in a July 2004 report by the UC Academic Council, a standing committee of the Academic Senate, which lauds system administrators for their handling of the issue and urges them to stay their present course.

The university, it says, "has acted in an exemplary fashion, working on its own and with consortia of other research universities to educate policymakers about the value of academic freedom and openness when possible and rejecting external funding when negotiation and education fails. … {UC] should continue to work with government agencies, on its own and through groups such as the AAU, to resist the blurring of the line between classified and unclassified research."

But Freedman, pointing to what she calls the Bush administration's "anti-intellectual" tilt, is not optimistic. Under previous presidents, she says, "We never spent any time worrying about clauses that said you have to be a U.S. citizen. That didn't happen. We never spent any time worrying about clauses that asked us to list our foreign students.

"If it's classified work, you want it classified," she continues. "You want to know upfront that this work is going to be used for military purposes and it's classified. But for basic, exploratory, we-don't-know-where-it's-going kind of work, it's the way the world moves on. And we're going to miss out on that…. Research isn't going to stop. It's just going to go somewhere else."

And what does the future hold? "It's probably going to get worse before it gets better," Freedman glumly predicts. "I just hope it doesn't get so much worse that we can't get good research done."