Last updated on

Threat Models Simplified (AI Edition)


Classic frameworks like STRIDE, DREAD and PASTA are widely used, but like any framework - there is an audience that needs it but isn’t big enough for the granularity. I have distilled these models into four simplified categories that are easier to apply in SIEM monitoring and modern AI-driven environments.


How to use this guide

This is not a checklist of things for your to check off. Its a template for a business exercise. The categories help you build towards

Four Simplified Categories

1. Identity and Access

Phrase: Who is in your system and what they can do
SIEM focus: Authentication logs, IAM events, PAM activity, VPN access, MCP access events, AI Crawl events

2. Data Integrity

Phrase: Ensuring data is not altered without authorization
SIEM focus: File integrity monitoring, database audit logs, Email access logs, source control activity, MCP actions taken

3. Confidentiality

Phrase: Keeping sensitive information from leaking
SIEM focus: DLP alerts, email security, cloud storage access logs, endpoint exfiltration patterns, LLM usage compared to authorized platforms,

4. Cybersecurity systems

Phrase: Systems stay online, actions remain traceable, and things that do alerts keep alerting SIEM focus: Firewall and IDS/IPS logs, system and application logs, service health events, audit trail completeness, EDR / Antimalware / Endpoint Security alerts, Service Disprutions


Why This Matters

By consolidating the core components of STRIDE, DREAD, and PASTA into four categories, you can assess both technical and business risk without the future governance debt of trying to align to whatever your CISO kicks out when you’ve 3x’d your organization. Incorporating AI into this categorization strengthens the summary, ensuring coverage of both traditional and emerging threats.

The original models

Threat ModelAcronymLetters & Meaning
STRIDESSpoofing identity
TTampering with data
RRepudiation (denying actions)
IInformation disclosure
DDenial of service
EElevation of privilege
---
DREADDDamage potential
RReproducibility of attack
EExploitability
AAffected users
DDiscoverability
---
PASTAPProcess for Attack Simulation
AAttack Simulation
SThreat Analysis
TThreat Analysis (continued emphasis in methodology)
AAlignment with business impact