Threat modelling with STRIDE

In short: Threat modelling systematically looks for threats before something is built or shipped. STRIDE is a mnemonic with six kinds of threat.

In more detail: You draw the system as a data flow diagram (actors, processes, stores, data flows, trust boundaries) and ask at every point: what can go wrong? Then you assess risks and plan countermeasures. Gaps found early are far cheaper to fix.

In Depth

• S – Spoofing: Pretending to be someone else. Countermeasure: authentication.
• T – Tampering: Altering data. Countermeasure: integrity, signatures.
• R – Repudiation: Denying an action. Countermeasure: logging, non-repudiation.
• I – Information disclosure: Exposing data. Countermeasure: encryption, access control.
• D – Denial of service: Taking a service down. Countermeasure: rate limits, redundancy.
• E – Elevation of privilege: Gaining more rights. Countermeasure: least privilege, input validation.
• Trust boundaries: Where data changes control (internet → server, server → database) most risks arise.
• Alternatives: PASTA, attack trees, LINDDUN (privacy).

See also: OWASP Top 10, Zero trust