Lawful Hacking and Continuing Vulnerabilities
The FBI’s legal battle with Apple is over, but the way it ended may not be good news for anyone.
Federal agents had been seeking to compel Apple to break the security of an iPhone 5c that had been used by one of the San Bernardino, Calif., terrorists. Apple had been fighting a court order to cooperate with the FBI, arguing that the authorities’ request was illegal and that creating a tool to break into the phone was itself harmful to the security of every iPhone user worldwide.
Last week, the FBI told the court it had learned of a possible way to break into the phone using a third party’s solution, without Apple’s help. On Monday, the agency dropped the case because the method worked. We don’t know who that third party is. We don’t know what the method is, or which iPhone models it applies to. Now it seems like we never will.
The FBI plans to classify this access method and to use it to break into other phones in other criminal investigations.
Compare this iPhone vulnerability with another, one that was made public on the same day the FBI said it might have found its own way into the San Bernardino phone. Researchers at Johns Hopkins University announced last week that they had found a significant vulnerability in the iMessage protocol. They disclosed the vulnerability to Apple in the fall, and last Monday, Apple released an updated version of its operating system that fixed the vulnerability. (That’s iOS 9.3you should download and install it right now.) The Hopkins team didn’t publish its findings until Apple’s patch was available, so devices could be updated to protect them from attacks using the researchers’ discovery.
This is how vulnerability research is supposed to work.
Vulnerabilities are found, fixed, then published. The entire security community is able to learn from the research, and—more important—everyone is more secure as a result of the work.
The FBI is doing the exact opposite. It has been given whatever vulnerability it used to get into the San Bernardino phone in secret, and it is keeping it secret. All of our iPhones remain vulnerable to this exploit. This includes the iPhones used by elected officials and federal workers and the phones used by people who protect our nation’s critical infrastructure and carry out other law enforcement duties, including lots of FBI agents.
This is the trade-off we have to consider: do we prioritize security over surveillance, or do we sacrifice security for surveillance?
The problem with computer vulnerabilities is that they’re general. There’s no such thing as a vulnerability that affects only one device. If it affects one copy of an application, operating system or piece of hardware, then it affects all identical copies. A vulnerability in Windows 10, for example, affects all of us who use Windows 10. And it can be used by anyone who knows it, be they the FBI, a gang of cyber criminals, the intelligence agency of another country—anyone.
And once a vulnerability is found, it can be used for attack—like the FBI is doing—or for defense, as in the Johns Hopkins example.
Over years of battling attackers and intruders, we’ve learned a lot about computer vulnerabilities. They’re plentiful: vulnerabilities are found and fixed in major systems all the time. They’re regularly discovered independently, by outsiders rather than by the original manufacturers or programmers. And once they’re discovered, word gets out. Today’s top-secret National Security Agency attack techniques become tomorrow’s PhD theses and the next day’s hacker tools.
The attack/defense trade-off is not new to the US government. They even have a process for deciding what to do when a vulnerability is discovered: whether they should be disclosed to improve all of our security, or kept secret to be used for offense. The White House claims that it prioritizes defense, and that general vulnerabilities in widely used computer systems are patched.
Whatever method the FBI used to get into the San Bernardino shooter’s iPhone is one such vulnerability. The FBI did the right thing by using an existing vulnerability rather than forcing Apple to create a new one, but it should be disclosed to Apple and patched immediately.
This case has always been more about the PR battle and potential legal precedent than about the particular phone. And while the legal dispute is over, there are other cases involving other encrypted devices in other courts across the country. But while there will always be a few computers—corporate servers, individual laptops or personal smartphones—that the FBI would like to break into, there are far more such devices that we need to be secure.
One of the most surprising things about this debate is the number of former national security officials who came out on Apple’s side. They understand that we are singularly vulnerable to cyberattack, and that our cyberdefense needs to be as strong as possible.
The FBI’s myopic focus on this one investigation is understandable, but in the long run, it’s damaging to our national security.
This essay previously appeared in the Washington Post, with a far too click-bait headline.
EDITED TO ADD: To be fair, the FBI probably doesn’t know what the vulnerability is. And I wonder how easy it would be for Apple to figure it out. Given that the FBI has to exhaust all avenues of access before demanding help from Apple, we can learn which models are vulnerable by watching which legal suits are abandoned now that the FBI knows about this method.
Matt Blaze makes excellent points about how the FBI should disclose the vulnerabilities it uses, in order to improve computer security. That was part of a New York Times “Room for Debate” on hackers helping the FBI.
Susan Landau’s excellent Congressional testimony on the topic.