Integrity

From a cybersecurity standpoint, integrity means that the data that you're receiving is complete, trustworthy and hasn't been altered. Obviously that's really important: when you're getting data, or when you're getting a document, or when you're receiving information, you want to make sure it's accurate.

The CIA triad and the pillars of cyber security

The CIA triad says we should be concerned about confidentiality, integrity and availability. Some would say that that's not enough, and that the pillars of cyber security should include authenticity and non-repudiation. Others may say, well, these are really closely related to integrity, and lump that into integrity. We're going to talk about what the difference is between those three.

Integrity, authenticity and non-repudiation

Let's use an example of an email that's being sent from Susan to David. Integrity says that that email has not changed along the way, so nothing has happened to this message; when David receives it, it's the original message that Susan sent. Authenticity is that this is from Susan — so when David receives this, it says that it's from Susan, and authenticity is that it is in fact from Susan. And non-repudiation is the fact that Susan can't deny sending it, and David can't deny receiving it.

Concerns

There are concerns over the integrity of messages when they're being sent. Traditionally email has not been an encrypted form of communication, which means anybody can take a look at that message, or even alter that message, between the sender, in this case Susan, and David. So at least traditional email, standard email, you should not consider as being a secure form of sending messages.

It's susceptible to things like an on-path attack. An on-path attack would be when somebody's sitting in between and can intercept that message for their own purpose, whether they're not sending it the other way, or maybe they're delaying the message from being sent, or perhaps they're altering the message as it comes in. It makes these messages susceptible to things like replay attacks, or just any kind of unauthorized information alteration — the alteration of the data — and then David gets a message that Susan didn't actually send.

It could come from outside the system too, so it's not just on-path attacks. Let's say somebody is impersonating, so they're outside the system, and they send a message to David pretending to be Susan, impersonating Susan. So we need some sort of integrity check or authenticity check to make sure that this mail is from who it says it's from, in this case Susan right here, and that it's not been altered.

Solutions

There are some solutions. It's pretty hard to fake video, although it's been getting easier to fake that. There are things like biometrics that could confirm somebody and who they are. Or a signature: we used to print things off, sign them, and then scan that and send that to a person in a PDF form. Or some sort of read receipt is available with some of these technologies, or you make somebody authenticate to see the message, and then we know that they received this.

All of these are kind of tricky solutions, because who wants to create a video, or how do you implement biometrics, or do you want to print something out to scan it, and the read receipts are kind of a problem. So we need better mechanisms, a digital way to be able to confirm these messages and the integrity of those messages.

Technology solutions to handle this: one is hashing, being able to use that fingerprint to confirm that nothing has been changed on this message. Or public key infrastructure — it uses public key cryptography to actually implement public key infrastructure. And then things like certificates, being able to use certificates to confirm somebody's identity, or digital signatures.