Trust models define how devices and systems verify and extend confidence to one another across a network, covering direct trust, third-party trust, hierarchy trust, web of trust, and how these frameworks underpin zero trust architecture.
Trust Models in Security
The most secure that I can make any network or any resource is to unplug them and never have any kind of connectivity. Of course, this isn't really reasonable in today's age; we need that level of connectivity. So at some point in time I'm going to need to extend some trust to different devices to make that level of connectivity. But how do I do that in a secure manner? There are different trust models.
There are a lot of different applications for these trust models. For instance, public key infrastructure is an example that we would apply some sort of trust model to. That is, we want to make it to different websites and we want a certain amount of verification that it's the actual website we're trying to reach, and so we do that with public key infrastructure using digital certificates and certificate authorities. Somehow we identify this trust that we have of a website, or the traffic that's getting to us, through a certificate authority, some other entity that we are going to trust.
One of the models is a direct trust. A direct trust is when there's a trust directly between two entities. In this case right here, we have a user that has a direct trust to the TechKnowSurge website. What that would look like is that the digital certificate would be installed directly onto the machine, and then any information that machine would receive from the TechKnowSurge website it would decrypt with that public key. As long as it's able to decrypt it with the public key, it knows that it was encrypted with a private key. So that's a direct trust.
Direct trusts don't really work on a large scale. That is, for every website we visit, we're not going to download and install a separate certificate. Those are extra steps that a user has to do.
So instead we developed something like a third-party trust, where a third party is going to be trusted to do the verification steps to trust something like the TechKnowSurge website. By us trusting this certificate authority, and them extending that trust now to the TechKnowSurge website, now users can go to the TechKnowSurge website and verify and trust this website right here. So that's third party: we're using a third party to do that trust.
Public key infrastructure also utilizes a hierarchy trust. There is some sort of top level trust that gets passed down to a subordinate, which also can get passed down again, and so it creates a hierarchy of trust, and from here we can create a lot of different trusts based off of the core of that.
A web of trust is quite a bit different. Let's say that each one of these is an entity. One of them happens to be me, and another one over here happens to be a stranger. I'm not going to immediately trust this stranger, but what I do have is trust with these other entities that are part of this web. Since I have a direct trust with the other entities, and these other entities have a direct trust with the stranger, then I'll create an indirect trust with this stranger, and there is a certain amount of trust there that I can have even though I haven't created a direct trust.
A trust anchor is not a model. It's actually an entity that we have a direct trust with that will pass on that trust to other entities. It's at the top, so it's the root. That is a trust anchor.
A chain of trust is not a model either. Instead, it's this concept of the link between all of the trusts. An example here is that we have the anchor of trust here at the top, and we also trust this entity down here. We trust this entity because of this chain of trust, so this would be the chain of trust for this entity that we trust right here.
One thing I did want to address here is this idea behind zero trust. It may sound like it's contradictory to these other trust models, but in fact it's actually very complementary to them, in that we actually use these different trust models to enforce zero trust.
The first thing you should probably understand is that zero trust is commonly associated with our internal network. That is, we are not going to trust a user that just plugs into our network and has physical access to our network. We're not going to just automatically assume that they should have access to the resources on that network. Instead, we're going to do some double checks before we give them access to resources on that network.
What we could do is we could actually set up things like a certificate authority on our local network, so this is actually a trusted source because we have set that up, and then we do a verification with this user before we give them access to the rest of those resources on that network. So it's still this never trust, always verify: we're actually verifying this user before they make some sort of connectivity to our resources. And we could actually use public key infrastructure, but just an internal public key infrastructure, to do those verification steps.
TechKnowSurge builds IT and cybersecurity professionals through hands-on, concept-first training built around real understanding — not memorization. Free interactive tools, structured programs, and 25+ years of real-world experience, all in one place.
Explore free tools and programs →