How the NSA Stole the Keys to Your Phone

(Editor’s Note: The following is a guest blog post from Julian Sanchez, Senior Fellow at the Cato Institute. It originally appeared on the Cato Institute website on February 20, 2015. The Cato Institute is a public policy research organization — a think tank – dedicated to the principles of individual liberty, limited government, free markets and peace).

A blockbuster story at The Intercept Thursday revealed that a joint team of hackers from the National Security Agency and its British counterpart, the Government Communications Headquarters (GCHQ), broke into the systems of one of the world’s largest manufacturers of cell phone SIM cards in order to steal the encryption keys that secure wireless communications for hundreds of mobile carriers—including companies like AT&T, T-Mobile, Verizon, and Sprint. To effect the heist, the agencies targeted employees of the Dutch company Gemalto, scouring e-mails and Facebook messages for information that would enable them to compromise the SIM manufacturer’s networks in order to make surreptitious copies of the keys before they were transmitted to the carriers. Many aspects of this ought to be extremely disturbing.

First, this is a concrete reminder that, as former NSA director Michael Hayden recently acknowledged, intelligence agencies don’t spy on “bad people”; they spy on “interesting people.” In this case, they spied extensively on law-abiding technicians employed by a law-abiding foreign corporation, then hacked that corporation in apparent violation of Dutch law. We know this was hardly a unique case—one NSA hacker boasted in Snowden documents disclosed nearly a year ago about “hunting sysadmins”—but it seems particularly poetic coming on the heels of the recent Sony hack, properly condemned by the U.S. government. Dutch legislators quoted in the story are outraged, as well they should be. Peaceful private citizens and companies in allied nations, engaged in no wrongdoing, should not have to worry that the United States is trying to break into their computers.

Second, indiscriminate theft of mobile encryption keys bypasses one of the few checks on government surveillance by enabling wiretaps without the assistance of mobile carriers. On the typical model for wiretaps, a government presents the carrier with some form of legal process specifying which accounts or lines are targeted for surveillance, and the company then provides those communications to the government. As the European telecom Vodaphone disclosed last summer, however, some governments insist on being granted “direct access” to the stream of communications so that they can conduct their wiretaps without going through the carrier. The latter architecture, of course, is far more susceptible to abuse, because it removes the only truly independent, nongovernmental layer of review from the collection process. A spy agency that wished to abuse its power under the former model—by conducting wiretaps without legal authority or inventing pretexts to target political opponents—would at least have to worry that lawyers or technicians at the telecommunications provider might detect something amiss. But any entity armed with mobile encryption keys effectively enjoys direct access: they can vacuum up cellular signals out of the air and listen to any or all of the calls they intercept, subject only to internal checks or safeguards.

There are, to be sure, times when going to the target’s carrier with legal process is not a viable option—because the company is outside the jurisdiction of the United States or our allies. Stealing phone keys in bulk is certainly a much easier solution to that problem than crafting interception strategies tailored to either the specific target or specific uncooperative foreign carriers. Unfortunately, the most convenient solution in this case is also a solution that gives the United States (or at least its intelligence community) a vested interest in the systematic insecurity of global communications infrastructure. We hear a great deal lately about the value of information sharing in cybersecurity: Well, here’s a case where NSA had information that the technology American citizens and companies rely on to protect their communications was not only vulnerable, but had in fact been compromised. Their mission is supposed to be to help us secure our communications networks—but having chosen the easy solution to the problem of conducting cellular wiretaps, their institutional incentives are to do just the opposite.

Finally, this is one more demonstration that proposals to require telecommunications providers and device manufacturers to build law enforcement backdoors in their products are a terrible, terrible idea. As security experts have rightly insisted all along, requiring companies to keep a repository of keys to unlock those backdoors makes the key repository itself a prime target for the most sophisticated attackers—like NSA and GCHQ. It would be both arrogant and foolhardy in the extreme to suppose that only “good” attackers will be successful in these efforts.

About the Author:

Julian Sanchez studies issues at the busy intersection of technology, privacy, and civil liberties, with a particular focus on national security and intelligence surveillance. In addition to his work as a Senior Fellow at the Cato Institute, he is a founding editor of the policy blog Just Security and contributing editor for Reason magazine. You can follow him on Twitter at @normative