Threat Modeling and AI Governance
This topic covers threat modeling frameworks, actor motivation, attack surface, abuse cases, STRIDE, ATT&CK, AI adoption risk, model threats, data exposure, and guardrails.
How to study for CompTIA SecurityX
Treat each SecurityX item as an enterprise security design decision: map requirements, model risk, place controls, engineer constraints, and operationalize evidence.
Core concepts
Concept 1
Threat modeling identifies actors, assets, trust boundaries, attack paths, and controls before or during system design.
Exam cue: Identify actor, asset, trust boundary, abuse case, and control.
Concept 2
AI adoption introduces governance, privacy, model, supply-chain, and excessive-agency risks.
Exam cue: Match framework to modeling objective and system maturity.
Concept 3
Security leaders should choose frameworks and controls that fit architecture, business context, and regulatory obligations.
Exam cue: Govern AI data, permissions, disclosure, and model-specific threats.
Risk pitfalls and guardrails
Threat modeling after deployment only as paperwork.
Guardrail: Avoid answers that ignore governance, overtrust the perimeter, deploy controls without placement logic, skip key management, or automate without guardrails.
Letting an AI assistant access sensitive systems without guardrails.
Guardrail: Avoid answers that ignore governance, overtrust the perimeter, deploy controls without placement logic, skip key management, or automate without guardrails.
Focusing on model output while ignoring training data and plugin risk.
Guardrail: Avoid answers that ignore governance, overtrust the perimeter, deploy controls without placement logic, skip key management, or automate without guardrails.
Memory anchors
STRIDE
STRIDE organizes threats as spoofing, tampering, repudiation, information disclosure, denial of service, and elevation of privilege.
Attack Tree
An attack tree decomposes paths an adversary could use to reach an objective.
Trust Boundary
A trust boundary marks where assumptions and controls change between components.
Abuse Case
An abuse case describes how a system could be misused.
ATT&CK
MITRE ATT&CK organizes adversary tactics and techniques for analysis and coverage.
CAPEC
CAPEC catalogs common attack patterns for threat analysis.
AI Guardrail
AI guardrails constrain model actions, data access, and behavior.
Prompt Injection
Prompt injection attempts to manipulate model behavior through crafted input.
Data Poisoning
Data poisoning corrupts training or reference data to influence behavior.
Model Inversion
Model inversion attempts to infer sensitive training information from model behavior.
Checkpoint rule
Do the check-up only after you can summarize each concept in one sentence and identify one dangerous pitfall from memory.
Knowledge Check (after reading)
Short check-up to confirm understanding of this module.
Check-up Questions
An architect is reviewing a new payment API before coding begins. The team needs to identify where identities, data, and trust assumptions change between the mobile app, gateway, services, and bank. What should the architect create first?
A threat-modeling workshop identifies that a service accepts a caller-supplied account identifier without verifying the caller owns the account. Under STRIDE, which threat is primary?
Answer all questions to submit.
Next step personalized recommendations
Continue learning
Move forward only after this module is stable.
What is Pass Harbor?
Completely free exam prep for 317 U.S. exams.
- Practice questions
- Flashcards
- Study guides
- Mock exams
- No registration
- No paywall
- Start instantly
“No more expensive exam prep. Quality study tools should be accessible to everyone.”