← All Tools

AI/ML Threat Model Builder

Describe your AI/ML system in 5 short steps; get the applicable threats from MITRE ATLAS, the OWASP Top 10 for LLM Applications (2025) and the OWASP ML Top 10, each with recommended mitigations and a documented risk rating.

Your architecture answers and the generated report stay in your browser (localStorage) — nothing is sent to any server. The threat model is sensitive, so it is produced entirely client-side.

Scope & limits. This is a structured aid that covers the listed, cited threat classes (MITRE ATLAS techniques, OWASP LLM Top 10 2025, OWASP ML Top 10, with NIST AI RMF where no ATLAS/OWASP item fits). It is not an exhaustive assessment and does not replace a full threat-modelling exercise or penetration test. Every threat shown names its source taxonomy; nothing here is invented.

How risk is scored & how to use this tool

Steps: (1) describe the model, (2) the data it handles, (3) what it can do, (4) who uses it and how it is integrated, (5) the controls already in place. Then Generate threat model to see the applicable threats, or use Try an example to pre-fill a sample.

Risk scale (reproducible): each threat has a base likelihood (1 Low / 2 Medium / 3 High) and impact (1–3). score = likelihood × impact; the rating is Critical for score ≥ 9, High for 6–8, Medium for 3–5, Low for 1–2. Your answers adjust the base: high-sensitivity data raises impact; no human oversight and missing controls raise likelihood; red-team testing lowers it. Each rating lists the factors it applied.

OWASP LLM Top 10 Checker → · MITRE ATLAS Search →

⚠ AI/ML Threat Model Builder

AI/ML Threat Model Builder