Let’s Have a Chat about Encryption

April 17, 2016

Discussions of encryption permeate tech and political news sites these days. All too frequently though, ideas presented in these discussions draw upon outdated analogies and confusion due to technological progression. I think it’s time to have a chat to answer some questions and move the discussion forwards.

A group of friends was walking home from dinner one day. After the usual discussions of gossip, work, and daily events had worn down, one of them broached a new topic.

“Have you guys been following the news related to the FBI wanting to force Apple to weaken the security of its devices?” asked James.

“I think that’s Apple’s view point,” responded Tim.

“Ah, so you have, then! What’s your opinion on the matter then?”

This time Emma interjected. “I think the government should stay out of Apple’s business.”

And other thought from the group: “I think when it comes to terrorists and other horrible criminals, Apple should have to unlock their phones if the FBI gets a warrant, but Apple shouldn’t have to make their devices less secure.”

“Yes and no,” Emma answered. “I believe there are times they should have to hand over information, but not provide the means to do that to the government.”

“That’s an interesting thought,” said James. “But I think I’m still a little hazy. Could you give me an example?”

Emma responded, “If the government can prove beyond a reasonable doubt that someone has committed a crime that threatens people’s safety—like what just happened—then Apple should hand over transcripts and a list of contacts. However, they shouldn’t have to hand over a guide to cracking Apple’s security procedures.”

James jumped back in again. “Well, now you two are kind of asking for two different things.

“The issue now is that those transcripts and lists of contacts might be encrypted. So are you saying that Apple should hand over the encrypted (meaning: useless without a key) transcripts and contacts, but that they shouldn’t be forced to withhold using encryption for transcripts and contacts in the first place?

“The reason why I’m asking is because Apple doesn’t have “security procedures” that can be cracked. It has encryption, and the question is whether people should be allowed to encrypt things.

“The whole point of encryption is to have security regardless of who you are. It’s founded in math; you can prove that a system is secure or insecure. You can’t both be able to unlock a phone at any time while also having a proof that no one can unlock the phone.”

“So what does that really mean? Couldn’t Apple just provide the information in an unencrypted form?” Emma wondered.

“Only if they have the key,” said James. “Apple’s argument is that they think holding the keys to someone’s encrypted secrets is a power too great for any one man or organization.”

The conversation died down, while everyone contemplated this position. Tim, who had been walking with his head tilted down facing the sidewalk, slowly looked up towards a flock of birds flying off in the distance.

“Your right to have your data encrypted and secure goes away when you kill innocent people,” Tim said, sternly.

The group paused once more to consider the implications of Tim’s statement. It James who broke the silence next.

“That’s a very rational viewpoint, but in saying that you’re also indicting the same innocent people you claim to defend. Encryption is blind; it it isn’t a privilege given to those worthy of it, nor is it reparations to be paid for a crime. Encryption is something everyone has access to. Not ‘should have access to’ or ‘gets to have access to.’ It’s free and open software.

“You can outlaw the ownership or use of that software, but it’s the same thing as outlawing the ownership of guns; the criminals will use guns or encryption anyway, and the general populace will not.

Tim was quick to respond this time. “Nothing in life is absolute. Read the Constitution: it’s life, liberty, pursuit of happiness, not subject to unreasonable search and seizure, etc. No one has an absolute right to encryption.

“Everyone should have access to encryption, but just like gun ownership, it can be forfeited, similar to how felons cannot own guns. If a court issues a warrant that a company help break encrypted data, they should comply.”

“What I’m trying to get at is that your viewpoint is contradictory,” James clarified. “Encryption cannot be forfeited after the fact, after it’s been determined that someone shouldn’t have access. If something is truly encrypted, then there is no force other than the key that one person owns that can unlock that data. I’m not claiming this as an opinion–that’s the technical definition of an encrypted system.

“So if we want companies to be able to comply with ‘breaking encrypted data,’ the data itself has to already be broken from the very beginning. It has to be insecure for everyone for it to be insecure for some one person. Without this, there’s no other way to retroactively ‘break encryption’ for a single person.

“Now in my opinion, a system that works by being able to be retroactively broken into is incredibly susceptible to misuse. This misuse could be either by governments that become tyrannical or by criminals who want to break into a system for their own gain. I personally feel that we as law-abiding citizens shouldn’t have to give up rights, dropping us down to the same levels as criminals, before we’ve even committed a crime.”

“And what I’m saying,” began Tim, “is this. Everyone is entitled to encrypt their data. With the proper warrants the government should be able to access someone’s personal data. If that data is encrypted, the government should be able to task really smart people to figure out the key for that one item. This should not affect someone else’s encryption.

“Now to the point of unbreakable electronic locks,” continued Tim. “If these are truly unbreakable, why aren’t they used everywhere and why are hackers able to break into things?”

James’ eyes shone for a moment before he started his response. “That’s actually a really good question. Let me back up a bit to explain.

“A ‘hacker’ in the traditional sense is actually not too much more than a con man. The majority of major breaks are what are called ‘social engineering attacks.’ Instead of exploiting vulnerabilities in hardware or software, they exploit vulnerabilities in people. From Wikipedia,

Social engineering, in the context of information security, refers to psychological manipulation of people into performing actions or divulging confidential information.

“So most hackers get passwords by tricking other people into divulging them. Obviously if you have the password, this whole discussion of encryption is moot. And because it’s so easy to get ahold of a password, there is a big push lately to use what’s called ‘two factor authentication.’ Instead of requiring just a password to unlock things, you need a password plus some sort of one-time-use code that is generated securely, usually through an app on your phone.

“The other major types of vulnerabilities access data through poor software coding practices. There are two common attacks: cross-site scripting attacks and SQL injection attacks. These attacks target the point of user-input. If not properly ‘sanitized’ to remove undesirable inputs, there are some circumstances where you can for example, enter computer code in a username field, and instead of just having that input be viewable on the screen, it accidentally gets treated as code and run on the hosted service. This is known as ‘remote code execution’: getting someone else to run code that you wrote. If you can achieve remote code execution, you can intercept what the software does after it has unencrypted the data.

“For relatively small applications, the attack vector is small; there are only so many surfaces in which to enter input, so it’s relatively feasible to check them all diligently. As software systems grow, complexities usually end up growing quadratically or even exponentially with the number of features added. It becomes really hard to exhaustively check all attack vectors, and usually large companies have whole departments devoted to this checking.

“So now that we know how the majority of ‘hacking’ works, let’s talk about encryption. From Wikipedia:

Encryption is the process of encoding messages or information in such a way that only authorized parties can read it. Encryption does not of itself prevent interception, but denies the message content to the interceptor.

“At this point, we can clarify the distinction between the two: ‘hacking’ is gaining unauthorized access to a software system, while ‘breaking encryption’ is reading a message that you weren’t intended to. Hacking deals with privilege escalation, and encryption deals with privacy.

“There are two types of encryption, but they both share a common goal: taking a hunk of data and scrambling it in a way that makes it easy to unscramble (given the password) yet near-impossible otherwise. The two types are called ‘encryption at rest’ and ‘encryption in transit.’ Encryption at rest deals with storing files securely on a hard drive, such that the privacy of those files are guaranteed even if the drive is stolen. Encryption in transit deals with communication: making sure that people can’t eavesdrop to discover your payment information, changed passwords, and other communications.

“To put this in context, let’s consider a simple messaging application. Using encryption in transit ensures that the messages sent from one person’s phone to the other can’t be eavesdropped, while encryption at rest ensures that the received messages stored on one person’s phone cannot be read unless the phone is unlocked with a password.”

Emma broke the monologue, asking, “So you’ve mentioned a lot about passwords and encryption, but here’s a simple question: does Apple have the ability to find out my password? Or if my password is highly secure, does that mean it’ll never be cracked in my lifetime?”

“This brings up an important issue: trust,” James responded. “Apple claims that it uses both of those forms of encryption, but there’s no way for you and me to verify that, other than take their word.

“To answer your question about passwords, the conventional wisdom would say no. Industry standard practice is to store only a ‘hash’ of your password, not the password itself. Every time you log in, the password you type is mangled using the same method that was used to mangle it when you first enter it, and these mangled versions are compared. This mangling process is ‘one-way’: you can’t take the mangled form and go back to the unmangled form, namely your password.

“But do you trust Apple to follow this method? Like we said: there’s no way to know other than their reputation. The only way to truly trust software is to verify the source code of it, finding out for yourself what it does. Unfortunately, Apple’s code is largely proprietary, so we can’t. However, there are many camps that advocate only using ‘open source software’: software whose source can be freely read and inspected. As you might imagine, this type of software is especially popular with software developers, as it’s fun to inspect the source code of the software you use, with the added benefit that you can place more trust in it.

“So, does Apple have the ability to find out your password? The answer is most likely no. They can reset your password if you ask for it to be reset, but this usually involves verifying your identity through your email provider.

“And regarding your second question, about whether your password can be cracked, you’re exactly right. A secure password can’t be cracked in any reasonable amount of time.

“There are a number of ways to attack password-based systems, the most common (other than social engineering attacks) being a ‘dictionary attack.’ This involves going through entire dictionaries, trying words, phrases, words with numbers in various places, and words with weird capitalization. So if you have a short, simple password, you’re likely vulnerable to a dictionary attack. Even if you have a short, complicated password, dictionary attacks tend to be fairly successful. That’s why it’s best to have long passwords, because as the dictionary crackers run out of words, it degenerates into a case of trying all possible strings. For example, if your password is 16 characters long, they’d have to try all 2616 possible inputs—we’re talking many, many guesses here.

“And when we typically talk about ‘encrypted’ data, we’re talking about data that has been locked up with a key that’s 128 characters long: 26128 possibilities! To put that in context, our best guess for the number of atoms in the universe is close to 1080. 26128 is an absolutely huge number.

“Since you sound curious, the math that protects these encryption schemes is a problem called ‘prime factorization.’ Basically, if you take two really big prime numbers p and q and multiply them, the number you get is pq, and only has two factors: p and q. If you know pq and want to get q, you ‘divide’ by the key p. Division on modern computers is super fast, so decrypting some data with the correct password is too. However, if you chose a large, random key p in the first place, it will take a really long time to guess what the key actually was before you can do the division.

“That’s a lot of information. With respect to the situation with Apple, all iPhone hard drives are encrypted at rest, meaning their data cannot be decoded without the lock password. The FBI wanted a way to circumvent this, by being able to try brute-forcing lock passwords without the phone self-destructing. In order to do this, Apple would have had to write code that allowed that to happen. Apple’s argument was that this act of writing code amounted to compelled speech, which is protected by the Constitution, and which has legal precedent when applied to code.

“After the case between FBI and Apple, it prompted many discussions about whether encryption itself should be legal. In particular, there is a draft of a bill in the Senate Intelligence Committee that is particularly foreboding for the legal use of encryption. That’s why I broached the subject in the first place, as I saw this in the news the other day. It’s pretty concerning that people want to see encryption weakened so that law-abiding citizens cannot have privacy for the sake of denying criminals their privacy.”

The group had finally arrived back home at their building. They passed inside and headed back upstairs. “No kidding that was a lot of information,” said Emma. “And I can definitely see where you’re coming from; it scares me a bit too when you put it that way. I’m glad you brought it up though. I feel like I have a grasp on things now that I didn’t before. Thanks.”


Jake on the Web

If you cared enough to read that far, you should consider following me on GitHub or paying a visit to my homepage. If this post was about one of my open source projects, make sure to star it on GitHub! I love hearing what people think, so feel free to open an issue or send me an email.

Haskell Development with Neovim

After a year and a half of using Haskell on and off, I’ve finally settled on a set of high-quality development and editor tools, using Stack and Neovim. Continue reading

Reach for Markdown, not LaTeX

Published on February 26, 2017

Troubleshooting Haskell Stack Setup on OS X

Published on August 03, 2016