We are in one of those phases again. The Paris attacks caused, once again, a cascade of demands for more surveillance and weakening of encryption. These demands appear every time, regardless of if the terrorists used encryption or not.
The perhaps most controversial demand is to make backdoors mandatory in communication software. Encryption technology can be practically unbreakable if implemented right. And the use of encryption has skyrocketed after the Snowden revelations. But encryption is not only used by terrorists. As a matter of fact, it’s one of the fundaments we are building our information society on. Protection against cybercrime, authentication of users, securing commerce, maintaining business secrets, protecting the lives of political dissidents, etc. etc. These are all critical functions that rely on encryption. So encryption is good, not bad. But as any good thing, it can be both used and misused.
And beside that. As people from the Americas prefer to express it: encryption is speech, referring to the First Amendment that grant people free speech. Both encryption technology and encrypted messages can be seen as information that people are free to exchange. Encryption technology is already out there and widely known. How on earth can anyone think that we could get this genie back in the bottle? Banning strongly encrypted messages would just harm ordinary citizens but not stopping terrorists from using secure communications, as they are known to disregard laws anyway. Banning encryption as an anti-terror measure would work just as well as simply banning terrorism. (* So can the pro-backdoor politicians really be that stupid and ignorant?
Well, that might not be the whole truth. But let’s first take a look at the big picture. What kind of tools do the surveillance agencies have to fight terrorism, or spy on their enemies or allies, or anybody else that happen to be of interest? The methods in their toolboxes can roughly be divided in three sections:
- Tapping the wire. Reading the content of communications this way is becoming futile thanks to extensive use of encryption, but traffic analysis can still reveal who’s communicating with whom. People with unusual traffic patterns may also get attention at this level, despite the encryption.
- Getting data from service provider’s systems. This usually reveals your network of contacts, and also the contents unless the service uses proper end-to-end encryption. This is where they want the backdoors.
- Putting spying tools on the suspects’ devices. This can reveal pretty much everything the suspect is doing. But it’s not a scalable method and they must know whom to target before this method can be used.
And their main objectives:
- Listen in to learn if a suspect really is planning an attack. This require access to message contents. This is where backdoors are supposed to help, according to the official story.
- Mapping contact networks starting from a suspect. This requires metadata from the service providers or traffic analysis on the cable.
- Finding suspects among all network users. This requires traffic analysis on the cable or data mining at the service providers’ end.
So forcing vendors to weaken end-to-end encryption would apparently make it easier to get message contents from the service providers. But as almost everyone understands, a program like this can never be water-tight. Even if the authorities could force companies like Apple, Google and WhatsApp to weaken security, others operating in another jurisdiction will always be able to provide secure solutions. And more skillful gangs could even use their own home-brewed encryption solutions. So what’s the point if we just weaken ordinary citizens’ security and let the criminals keep using strong cryptography? Actually, this is the real goal, even if it isn’t obvious at first.
Separating the interesting targets from the mass is the real goal in this effort. Strong crypto is in itself not the intelligence agencies’ main threat. It’s the trend that makes strong crypto a default in widely used communication apps. This makes it harder to identify the suspects in the first place as they can use the same tools and look no different from ordinary citizens.
Backdoors in the commonly used communication apps would however drive the primary targets towards more secure, or even customized, solutions. These solutions would of course not disappear. But the use of them would not be mainstream, and function as a signal that someone has a need for stronger security. This signal is the main benefit of a mandatory backdoor program.
But it is still not worth it, the price is far too high. Real-world metaphors are often a good way to describe IT issues. Imagine a society where the norm is to leave your home door unlocked. The police is walking around and checking all doors. They may peek inside to check what you are up to. And those with a locked door must have something to hide and are automatically suspects. Does this feel right? Would you like to live in a society like that? This is the IT-society some agencies and politicians want.
(* Yes, demanding backdoors and banning cryptography is not the same thing. But a backdoor is always a deliberate fault that makes an encryption system weaker. So it’s fair to say that demanding backdoors is equal to banning correctly implemented encryption.