Books have been written about President Eisenhower’s famous farewell warning in 1961 about the “military-industrial complex,” and what he described as its “unwarranted influence.” But an even greater leviathan today, one that the public knows little about, is the “intelligence-industrial complex.”
The saga of the private sector’s involvement in the NSA’s scheme for permanent mass surveillance is long, complex, and sometimes contentious. Often, in ways that appeared to apply indirect pressure on industry, the NSA has demanded, and received, approval authority—veto power, basically—over telecom mergers and the lifting of export controls on software. The tech industry, in more than a decade of working-group meetings, has hashed out an understanding with the intelligence community over greater NSA access to their systems, including the nation’s major servers (although it is not yet clear to what degree the agency had direct access). “I never saw [the NSA] come and say, ‘We’ll do this if you do that,’ ” says Rebecca Gould, the former vice president for public policy at Dell. “But the National Security Agency always reached out to companies, bringing them in. There are working groups going on as we speak.”
Indeed, the cooperation was usually “voluntary” in large part because companies couldn’t afford to seem uncooperative, says another private-sector official who would speak about classified issues only on condition of anonymity. “The ways that pressure works in Washington are very subtle,” he says. “No one’s getting bribed, or punished outright. But it’s the good little Indian that gets rewarded. And these companies needed the goodwill of the NSA and other agencies.”
...
The origins of the intelligence-industrial complex date back to World War II and a program called Shamrock, under which the NSA came to an agreement with ITT and other companies to collect outgoing telegrams and international cables. That secret program was exposed in the 1970s, in an earlier incarnation of the current scandal, and helped lead to the famous Church-Pike congressional hearings on intelligence abuses (which in turn led to the FISA law).
But the latest chapter in the saga, involving Silicon Valley, begins in the immediate aftermath of the Cold War, when Hayden and other senior NSA officials, including his predecessor, Ken Minihan, were in a state of near-panic. Not only had the Soviet Union—the chief object of the NSA’s spying, and its raison d’etre—disappeared from the map, but now the agency also realized that the main threat was going to be “super-empowered” individuals—terrorists—who might be talking on cell phones or computers anywhere on earth. Above all, these new bad guys were using private technology, rather than the sort of intra-government communications systems that the NSA used to monitor in the Soviet Union or China. Not by coincidence, during the Cold War, the NSA often had the biggest hand in designing its own detection equipment. “We were America’s Information Age enterprise during America’s Industrial Age. We had the habit of saying, ‘If we need it, we’re going to have to build it,’ ” Hayden says. “But in the outside world, there was a technological explosion in the two universes that had been at the birth of the agency almost uniquely ours: telecommunications and computers. The Internet began as a combination of those two—you could probably draw a good history as to what we did to create the American computing industry back in the ’50s.”
Yet once that computing industry took off in Silicon Valley, to be followed by the rise of Internet technology and “smart” phones, the NSA found itself left further and further behind, never to catch up. In a period of a decade or so, Hayden said, the agency went “from chasing the telecommunications structure of a slow-moving, technologically inferior, resource-poor nation-state—and we could do that pretty well—to chasing a communications structure in which an al-Qaida member can go into a storefront in Istanbul and buy for $100 a communications device that is absolutely cutting-edge.” And he could then contact other terrorists in every country, particularly in the United States.
The NSA’s early response was to try to barge its way back into the domestic-surveillance business with devices such as the “Clipper” chip, an encryption tool developed by the agency that it wanted telecommunications companies to adopt. Consumer-protection and computer-privacy groups howled in protest, and industry resisted the government telling it how to manage its technology. The idea was dropped after a few years in 1996.
Then, in the late 1990s, a furor erupted over export controls on software encryption. The NSA sought to bar exports of the best encryption technology, fearing what would happen if enemies got hold of it. As it had done with the Clipper chip, Silicon Valley countered that by holding the tech sector back, the government was hurting U.S. national security. It argued that the U.S. would fall far behind other nations in a critical industry unless those controls were lifted.
After months of battles, a quiet quid pro quo was struck, according to a former senior intelligence official: We’ll let you export first-rank encryption, the government said, but we want to get a first look at what you’re developing and a back door into it. A Clipper chip wasn’t needed, after all, if the government was going to get access to servers and telecom data. “The way the encryption deal was worked out was that, in the end, controls were liberalized in various stages, in 1997, ’98, and ’99, and all of the liberalizations had a single bottom line: All products had to be reviewed by the NSA,” says William Reinsch, who was undersecretary of Commerce during a critical period in the 1990s when the NSA was undergoing a dramatic decline from the chief innovator of America’s spying technologies, and instead finding itself falling behind Silicon Valley and the telecom industry. “That review meant [NSA] got to look at them.… It was a source of considerable irritation to companies—not the basic fact of it but that NSA wanted to continue to do it for every product.
The NSA came to understand that it was better for them if the world was fully populated with technologies that they knew and understood.” Especially if the agency had the consent of industry to penetrate those technologies. If industry refused, the NSA had the unique ability to both reward and punish, thanks to its implicit veto power over deals and exports, Reinsch says. Though the public didn’t know it, the agency also became a major presence when the nation’s telecom industry went through a revolution, moving from the Bell system to a flurry of start-ups and a blizzard of mergers. “The NSA’s ability to access [telecommunications data] became a factor in all those telecom acquisitions,” Reinsch says. It and other intelligence agencies “weighed in and said we want to review this transaction. We want to say no if we think it’s a bad idea.” The NSA rarely exercised that right, but its leverage was useful in co-opting the tech and telecom sectors into its plans.