CareerExamPro

EC-Council AI Security (C|OASP & C|RAGE) & Security+ V8 (SY0-801) practice test

9 practice questions for this exam, each with a full explanation and cited rule or blueprint reference.

Start practising

SecAI+, SecOT+, AAISM/AAIA & CMMC: practice drill

10 realistic exam questions, graded as you go, with the exact rule or statute citation for every answer. Start free, no sign-up.

Exam Snapshot & Official Blueprint

Official Exam Fee
2026 Public Beta / GA Exam Fee
Format & Time Limit
Scenario MCQs · Offensive AI Security & Responsible AI Ethics
Passing Standard
Blueprint-aligned passing cut score
Typical Prep Incumbents
2026 public beta includes zero official study materials

Official Domain Breakdown

Exam DomainWeight
Offensive AI Red-Teaming, Multi-Turn Jailbreaks & Agent Exploitation (C|OASP)45%
Responsible AI Governance, Fairness & Ethics Assurance (C|RAGE)30%
Modern AI-Driven Cyber Defense & Zero Trust (Security+ SY0-801)25%

High-Yield Formulas, Statutes & Rules Cheat Sheet

Frequently Asked Questions

Why does EC-Council's 2026 AI beta have no study material?

During public beta testing, certifying bodies intentionally withhold official textbooks to validate item difficulty against real-world practitioner knowledge.

Practice tests that prepare you for this exam

Practise by topic

About this exam

Agency / Issuer
EC-Council / CompTIA
Region
Global / Online

Exact question counts, passing scores, and allowed reference materials vary by testing window and state/agency. Always confirm current rules with the official testing provider.

Sample questions

A SOC deploys an ML malware classifier trained on PE header attributes and Windows API import tables. Without modifying the training pipeline or model weights, an adversary iteratively appends benign overlay bytes and unused legitimate DLL imports to ransomware samples until the classifier scores them as benign. Under MITRE ATLAS, how is this attack classified, and what strengthens defense?

  1. Model supply-chain compromise via serialized weights; strengthen defense by migrating model checkpoints from pickle files to cryptographically signed SafeTensors.
  2. Data poisoning via label flipping in the SIEM lake; strengthen defense by freezing training datasets and requiring two-person approval for label edits.
  3. Model extraction via surrogate querying; strengthen defense by rounding confidence probabilities to coarse tiers and throttling repeated API requests per client.
  4. Evasion via adversarial perturbation at inference time; strengthen defense with adversarial training, ensemble models, and behavioral runtime telemetry.
Show the answer

D. Evasion via adversarial perturbation at inference time; strengthen defense with adversarial training, ensemble models, and behavioral runtime telemetry.

Manipulating input features at inference time to force misclassification without altering training data or model weights is an evasion attack (MITRE ATLAS). Combining adversarial training on perturbed binaries with runtime behavioral EDR detections reduces reliance on brittle static PE features.

Reference: CompTIA SecAI+ (CY0-001) / MITRE ATLAS (ML Attack Staging & Evasion)

An autonomous DevOps AI agent runs with a cloud-administrator service token to clean up stale staging environments. After misreading an ambiguous change ticket, it deletes a production database snapshot vault. Under OWASP Top 10 for LLM Applications and Agentic AI guidance, which vulnerability class and architectural fix directly apply?

  1. Excessive Agency; enforce least-privilege tool scopes, separate read/write tokens, and require human confirmation for destructive actions.
  2. Unbounded Consumption; enforce per-session token budgets, rate limits, and automated billing circuit breakers.
  3. System Prompt Leakage; move secret cloud credentials out of the system prompt template and inject them via encrypted environment variables at container startup.
  4. Sensitive Information Disclosure; apply regex output masking to strip AWS account identifiers and database table names from agent operational audit logs.
Show the answer

A. Excessive Agency; enforce least-privilege tool scopes, separate read/write tokens, and require human confirmation for destructive actions.

Granting an AI agent overly broad permissions, unnecessary tools, or unchecked autonomy to execute destructive state changes is Excessive Agency (OWASP LLM06). Least-privilege scoped tool tokens, environment isolation, and human approval for high-impact actions contain blast radius.

Reference: CompTIA SecAI+ (CY0-001) / OWASP Top 10 for LLM Applications (LLM06: Excessive Agency)

While auditing an automated resume-screening system, an AAIA auditor confirms that explicit gender and ZIP-code fields were removed from training data, yet selection rates still show a 26% adverse disparity against a protected group due to correlated proxy features such as athletic clubs and graduation-year gaps. What should the assurance report conclude?

  1. Because protected-class columns were excluded from the input schema, the model is statistically unbiased and requires no additional fairness controls.
  2. Fairness through unawareness failed via proxy features; mandate subgroup disparate-impact testing, explainability reviews, and bias mitigations.
  3. Replace the classifier with an uncalibrated black-box LLM prompt so individual feature weights can no longer be traced back to candidate attributes.
  4. Re-inject protected-class columns as unmonitored input features without altering the training loss function so the algorithm can self-correct automatically.
Show the answer

B. Fairness through unawareness failed via proxy features; mandate subgroup disparate-impact testing, explainability reviews, and bias mitigations.

Simply omitting explicit protected attributes ('fairness through unawareness') fails when correlated proxy variables reconstruct group membership. Defensible assurance requires subgroup disparate-impact measurement, explainability analysis of proxy features, and targeted fairness mitigation controls.

Reference: ISACA AAIA & EC-Council C|RAGE / NIST SP 1270 (Managing Bias in AI)

Taking this exam soon?

Tell us which exam domains or calculation drills you want more practice on.

Other 2026 AI Security, OT & CMMC exams