STRIDE threat modeling explained with a simple login example

STRIDE is a checklist of six threat types used during threat modeling: Spoofing, Tampering, Repudiation, Information disclosure, Denial of service and Elevation of privilege. Each one violates a security property, so the list tells you what to ask about every component and data flow in your diagram.

The six types and what they violate

OWASP's threat modeling process pairs them like this:

• Spoofing: Authentication
• Tampering: Integrity
• Repudiation: Non-repudiation
• Information disclosure: Confidentiality
• Denial of service: Availability
• Elevation of privilege: Authorization

A login flow example

Imagine a browser, a web app, an API and a user database. Apply each type to the login request.

• Spoofing: Can someone pose as another user? Think of stolen passwords, weak reset flows or guessable session IDs.
• Tampering: Can a request or a stored value be changed in transit or at rest? Think of an unsigned token or a role field the client controls.
• Repudiation: Could a user deny an action because you have no reliable log? Check that sign-ins and privilege changes are recorded.
• Information disclosure: Do error messages reveal whether an account exists? Are credentials logged or sent without TLS?
• Denial of service: Can an attacker lock out real users or exhaust the login endpoint?
• Elevation of privilege: Can a normal user reach an admin function by changing a parameter?

How to use it in a session

1. Draw the data flow diagram.
2. For each element, run through the six types as prompts.
3. Write each plausible threat as a sentence.
4. Pick a response and an owner. OWASP lists accept, eliminate, mitigate or transfer.

You do not need to find every threat. You need to find the ones worth fixing while the design is still changeable.

Limits of STRIDE

It is a prompt list, not a method for ranking risk. It can overlook business logic abuse and supply chain issues unless you add them as elements in the diagram. Use it with the four questions approach. See threat modeling in four questions.

Where Safeguard fits

Supply chain threats such as a compromised dependency show up as tampering in your diagram. Safeguard's SBOM and dependency scanning gives you an inventory to put in that diagram.

Turning threats into work

A list of threats helps only if it becomes tickets. Write each threat in one line with this pattern: who does what to which asset, and what happens. For example, "An attacker replays a captured reset link and takes over the account." Then add a response, an owner and a due date.

Rank by asking two plain questions: how easy is it, and how bad would it be. Fix the easy and severe ones first. For the rest, record why you accepted them. Future you will be glad to see the reasoning when the system changes.

Quick check

Sources

• OWASP Threat Modeling Process

Last reviewed: 19 October 2026.