10 realistic exam questions, graded as you go, with the exact rule or statute citation for every answer. Start free, no sign-up.
Timed mock exam
15 questions in 18 minutes, drawn from every topic. No answers shown until the end, then a full score breakdown, your weakest topics and a review of what you missed.
Prices are in US dollars, plus tax where it applies. Stripe, our payment processor, calculates tax at checkout.
All sales are final. If something is really wrong, such as a duplicate charge or a purchase that does not unlock what you bought, write to support@careerexampro.com and we will put it right.
One-time payment. Nothing renews.
You have an exam ticket for this practice test. Start the timed mock exam to use it.
Exam Snapshot & Official Blueprint
Official Exam Fee
$404–$425 (Pearson VUE / OnVUE — Launched Feb 2026)
Format & Time Limit
Max 60 questions (MCQ + PBQ) · 60 minutes
Passing Standard
600 out of 900 scaled score
Typical Prep Incumbents
Brand-new Feb 2026 launch — very few third-party Q-banks exist yet
Official Domain Breakdown
Exam Domain
Weight
Securing AI/ML Models, Data Pipelines & DevSecOps (MLSecOps)
30%
Defending Against Adversarial, Evasion & Prompt-Injection Attacks
28%
AI Governance, Risk & Global Compliance (EU AI Act / NIST AI RMF)
OWASP LLM Top 10 & MITRE ATLAS: Distinguish Direct vs. Indirect Prompt Injection (hidden inside fetched web pages/RAG docs), Training Data Poisoning, Model Inversion/Extraction, and Excessive Agency in tool-calling agents.
RAG Access-Control Leakage: Always enforce user-level ACL metadata filtering inside the vector store BEFORE context enters the LLM prompt—never rely on prompt instructions alone to hide confidential chunks.
AI-Assisted SOC Guardrails: Validate AI-summarized alerts and generated IOCs against raw SIEM telemetry to prevent hallucinated hashes/IPs or attacker-injected log payloads from triggering automated SOAR lockouts.
Exam Pacing (60 Qs in 60 min): Flag complex performance-based scenarios on pass 1 and lock in concise MCQs at ~45 seconds each.
Frequently Asked Questions
When did CompTIA SecAI+ (CY0-001) launch?
CompTIA launched SecAI+ (CY0-001) in February 2026 as the first vendor-neutral technical certification for securing AI systems and AI-assisted security operations.
Are there prerequisites for CompTIA SecAI+?
No mandatory prerequisite, though CompTIA recommends 3–4 years of IT experience and 2+ years of cybersecurity (Security+ / CySA+ level).
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?
Model supply-chain compromise via serialized weights; strengthen defense by migrating model checkpoints from pickle files to cryptographically signed SafeTensors.
Data poisoning via label flipping in the SIEM lake; strengthen defense by freezing training datasets and requiring two-person approval for label edits.
Model extraction via surrogate querying; strengthen defense by rounding confidence probabilities to coarse tiers and throttling repeated API requests per client.
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.
API telemetry shows an external B2B tenant sending 1.8 million systematically spaced synthetic feature vectors to a proprietary fraud-scoring endpoint over 48 hours, recording full floating-point class probabilities for every response. Which AI threat is most likely in progress, and which API control reduces risk without breaking legitimate customer scoring?
Indirect prompt injection through transaction memos; mitigate by stripping special characters from string fields and adding a static system prompt wrapper.
Vector database embedding inversion; mitigate by rotating index encryption keys and migrating embeddings from cosine similarity to Euclidean distance.
Model extraction to build a surrogate replica; return coarse labels or rounded scores, apply rate limits, and flag query anomalies.
Concept drift caused by seasonal spending shifts; mitigate by scheduling automated nightly model retraining on the incoming synthetic API payloads.
Show the answer
C. Model extraction to build a surrogate replica; return coarse labels or rounded scores, apply rate limits, and flag query anomalies.
High-volume, systematically distributed probing combined with collecting exact floating-point confidence vectors enables model extraction (stealing a functionally equivalent surrogate model). Truncating precision to coarse bands or binary decisions, throttling tenants, and detecting abnormal query distributions significantly slow extraction.
An MLOps CI/CD runner pulls third-party open-weights checkpoints in legacy PyTorch .pkl/.pt format from a community model hub and deserializes them directly inside a GPU container attached to a broad production IAM role. What is the primary supply-chain vulnerability, and what MLSecOps standard prevents exploitation?
Arbitrary code execution via pickle deserialization; mandate SafeTensors, verify signatures, scan artifacts, and load in least-privilege sandboxes.
GPU side-channel memory leakage across virtual machines; mandate single-tenant bare-metal instances and disable CUDA shared-memory caching inside all build runners.
Gradual weight quantization loss across container rebuilds; mandate FP32 precision exports and archive intermediate optimizer checkpoints in immutable object storage.
Unintended hallucination of deprecated library APIs; mandate deterministic decoding with temperature zero and enable automated unit-test retries in CI/CD.
Show the answer
A. Arbitrary code execution via pickle deserialization; mandate SafeTensors, verify signatures, scan artifacts, and load in least-privilege sandboxes.
Python pickle files execute embedded deserialization routines (such as __reduce__) upon loading, allowing malicious model weights to run arbitrary shell commands with the container's IAM privileges. Enforcing non-executable formats like SafeTensors, verifying cryptographic hashes/signatures, and isolating model loaders with least privilege eliminates this attack path.
Reference: CompTIA SecAI+ (CY0-001) / OWASP Top 10 for LLM Applications (LLM03: Supply Chain)
Taking this exam soon?
Tell us which exam domains or calculation drills you want more practice on.