Does Encryption Really Protect Terrorists?
In the past few months, we’ve seen several proposals from various governments to require companies to put some sort of backdoor into their encryption, so that the government can catch terrorists. In the wake of the Paris bombings, calls for a backdoor in encryption have been renewed.
What does it mean to put a “backdoor” in encryption technology? I think it’s best to work by analogy.
Imagine terrorists are hiding behind physical locked doors. Perhaps we could solve this problem if every lock manufacturer in this country was required to send a copy of your keys (or a separate master key) to the government for every lock sold in this country? Don’t worry, the government will only use these keys for good. And they will never let these keys fall into the wrong hands. Some petty disgruntled employee who works in the extensive bureaucracy created for this key storage and distribution can never happen, honest. After all, Snowden was a one-shot that’ll never happen again (most likely because the next thief won’t be a whistleblower and we won’t even know anything was stolen). Also, that large number of NSA workers who used wiretaps for their personal fun and profit would magically not exist in this key storage situation.
OK, so lets scrap the key storage idea. What if locksmiths in our country were not allowed to sell the best locks in the world, but instead only locks with a design flaw sanctioned by the government? The government would know the secret method needed to pick this lock, but no one else would. Don’t worry, the government will keep this secret safe. Honest. In this universe too, all of the employees involved in designing this lock, all of the manufacturers who make this lock, and all the locksmiths who service the locks will all promise to play nice and never ever tell anyone what this design flaw is. They’ll take blood-oaths or spit-handshakes or something. Also, it will be a magical design flaw that can not be reverse-engineered by people who are holding an actual lock in their hands. Perhaps the government will make it out of that very expensive element canttakeapartanium.
Yeah, that won’t really work either.
But what if it did work? What if we did manage to get some sort of door lock scheme in place where the government really did keep the backdoor secret safe, and only ever used it for good?
Umm… the terrorists would get their locks from some other country.
So we gain nothing, even if the backdoor lock was not a pipe dream. But it probably is a pipe dream, so the entire world would know that our country has the worst door locks in the world. And by door locks here, I obviously mean encryption. Privacy. Security. The technology that keeps all your money secure in the bank. And which keeps your identity from being stolen. And which, if used properly, would keep your email from being leaked to the press after you make a “funny” movie about the assassination of Kim Jong Un and he gets pissed off and hires hackers to break in and embarrass everybody. Allegedly.
So the world will know that we have the worst cryptography. All the hackers and terrorists and countries and mobsters on the entire globe will double down on their investment into hacking into our country’s infrastructure. Heck, they’ll probably triple-down or quadruple-down or more. It’s just too tempting. It’s just like those nature documentaries where the cheetah looks for the easy prey.
Let’s make sure you’re keeping score here. We spend billions of dollars, weakening the security of our own country, putting all of our privacy, our money, our commercial secrets, our government secrets at risk, and we never achieve the benefit that we were supposed to get. And in fact, by weakening our own security, we make ourselves into a giant target to every country, terrorist, criminal, and bored hacker on the entire planet. All in the service of accomplishing nothing good.
Let’s try one more analogy that’s a bit closer to the mark: phone taps. The government has been tapping people’s phones for over a century. When phones were new, and rules about wiretaps were created we lived in an interesting era: an era of nearly perfect privacy. That’s not a joke — circuit switched telephone networks were only vulnerable at the endpoints. The middle was an incomprehensible tangle of wires. It was impossible to monitor calls from the middle. This is where we get that TV trope about keeping people on the line while a call was being traced — with circuit switched phone calls, a bunch of engineers would run around from one physical relay to the next until they had traced every physical wire involved in a phone call. This could take ten minutes, twenty minutes, even longer for long-distance calls where you have to coordinate engineers chasing relays in multiple locations across the country. This was just tracing a call, not bugging it. You just couldn’t wiretap from a central location. Bugging a phone at least involved climbing up a phone pole next to the house, and sometimes involved entering the premises. You had to be at the endpoint or it was impossible.
Encryption is no worse than this. You can’t decrypt the data in the middle, but you can still spy on the endpoint. You can watch the traffic being decrypted. You can steal the encryption keys from the endpoint after obtaining a suitable warrant. The claim by proponents of a backdoor is that encryption is somehow new, that it allows terrorists to hide in a new way that didn’t exist before. The reality is that the only new thing is the government’s recent and short-lived ability to do wholesale wiretapping from a central location.
Here’s what it comes down to. Governments are looking for a magic wand to help them catch terrorists. Or a tooth fairy. Better yet, a terrorist fairy. They want to sit in a central location and have the terrorist fairy deliver a daily list of names to them. The reality is there is no such thing as a terrorist fairy. No magical being can rescue us. Nor can magical technology. Police have to do real police work, just like they always have.
Somebody needs to tell the government that magic wands and terrorist fairies don’t exist.
To answer the original question: of course encryption protects terrorists. Because encryption protects everyone without differentiation. Encryption can be considered a synonym for “protection”, in which case the title question is “does protection protect terrorists?” Yes. Duh. But this formulation makes it clearer: weakening all private protection so that terrorists won’t be protected is clearly insane. We need protection for ourselves, and we need it to be the best available. There is no benefit in having weaker protection than the bad guys have. Now can we please do real work stopping terrorist and stop wishing for terrorist fairies?