ISTQB Certified Tester AI Testing

CT-AI · ● Active · Professional · ISTQB

Specialist extension module for testing AI-based systems, machine learning models, and generative AI. Prerequisite: ISTQB Certified Tester Foundation Level (CTFL). Current version: v2.0 (released April 2026).

Exam facts

Vendor source — ISTQB Certifications ↗
Official exam guide — CT-AI Syllabus v2.0 ↗
Exam objectives — CT-AI Sample Questions ↗
Exam registration — ASTQB Exams ↗

About

The ISTQB Certified Tester AI Testing (CT-AI) is a specialist-level certification validating expertise in testing AI-based systems, including machine learning models, neural networks, and generative AI applications. Launched in 2021 and updated to v2.0 in April 2026, it serves testers, test analysts, QA engineers, and developers transitioning to AI-driven quality assurance. The v2.0 refresh adds comprehensive coverage of generative AI and large language models (LLMs), strengthens statistical foundations for ML testing, and emphasizes deployment and production monitoring practices. CT-AI is vendor-neutral and internationally recognized.

Why CT-AI matters: As AI/ML systems enter production across fintech, healthcare, autonomous vehicles, and enterprise software, testing these systems requires specialized knowledge distinct from traditional QA. CT-AI addresses this gap by providing standardized, internationally recognized credentials for professionals who can assess AI model quality, detect bias, validate performance metrics, and ensure safe deployment. The certification bridges QA and data science, preparing testers to collaborate effectively with data scientists and ML engineers.

Domain context — Quality Assurance

Professional-level QA testing specialization for artificial intelligence and machine learning systems. Part of the broader ISTQB Testing Ecosystem covering foundation, advanced, and specialist certifications.

Read full deep dive — ISTQB Testing Ecosystem → (file not yet created)

Topics covered

The CT-AI v2.0 syllabus covers testing across the complete AI/ML lifecycle. Version 2.0 (released April 2026) removes content on "using AI for testing" and focuses entirely on "testing AI-based systems."

• Fundamentals of AI Testing — Core concepts, AI/ML paradigms, testing principles specific to intelligent systems, differences from traditional software testing
• Machine Learning Fundamentals — Supervised learning (classification, regression), unsupervised learning (clustering), reinforcement learning; model types, architectures, and workflows
• Data & Input Testing — Bias detection, disparate impact analysis, data representativeness, label quality assurance, handling class imbalance, data pipeline validation, feature engineering validation
• Machine Learning Model Testing — ML-specific test strategies, performance metrics (accuracy, precision, recall, F1-score, AUC-ROC), test levels, adversarial testing, metamorphic testing, robustness evaluation
• Neural Networks — Perceptron implementation and hands-on exercises, deep learning fundamentals, layer-wise testing, activation functions, coverage measures (neuron coverage, k-neuron coverage, boundary testing)
• Generative AI & Large Language Models — Testing LLMs, prompt engineering validation, output quality assessment, red teaming techniques, detecting hallucinations, jailbreak testing, safety evaluation
• Statistical Testing for ML — Hypothesis testing, statistical significance, confidence intervals, power analysis, avoiding p-hacking in model evaluation
• Deployment & Production Monitoring — Pre-deployment testing, concept drift detection, data drift detection, model performance monitoring, A/B testing, shadow deployment, model retraining triggers
• Quality Characteristics for AI — Reliability, robustness, fairness/bias, explainability (XAI), safety, security, privacy considerations in AI systems
• Test Tools & Automation — ML testing frameworks, Python testing libraries, data validation tools, ML pipeline orchestration, continuous validation

Source: ISTQB CT-AI v2.0 Syllabus Release ↗

Common skills at Quality Assurance · Professional

Shared professional-level testing skills across the QA domain — not specific to AI testing.

• Test strategy and test planning for complex systems
• Risk-based testing and test prioritization
• Advanced test case design and optimization
• Test automation frameworks and tools (Selenium, TestNG, Pytest)
• Regression and continuous integration testing
• Performance and load testing methodologies
• Root cause analysis and defect management
• Test metrics, reporting, and KPI tracking
• Requirements analysis and test scope definition
• Cross-functional collaboration with development and product teams
• Traceability matrix maintenance and coverage analysis
• Agile and DevOps testing practices
• Test data management and security considerations

Recommended courses at Quality Assurance · Professional

Course selection notes: Official ISTQB training partners provide accredited 3-day instructor-led courses. Udemy courses are self-paced and best for self-study. All listed courses address CT-AI v2.0 content.

Practice exams

Books

Book note: The Itkin et al. guide is the authoritative companion to the official CT-AI v2.0 syllabus, written by practitioners from Exactpro Labs. Includes structured chapter questions and a full mock exam.

Typical job titles at Quality Assurance · Professional

AI Test Engineer · Machine Learning Quality Assurance Engineer · Data Quality Analyst · AI Testing Specialist · Test Analyst (AI Systems) · Quality Assurance Engineer (ML/AI) · Test Automation Engineer (AI Systems) · AI Systems Tester

(Job titles drawn from current postings on ASTQB, LinkedIn, and ISTQB partner job boards that list CT-AI as required or preferred.)

Salary

Salary context: Ranges reflect QA professionals with AI/ML testing expertise. Certification alone does not guarantee salary; experience and industry (fintech, healthcare, autonomous systems) significantly impact compensation. AI-focused roles typically command 15–25% premium over general QA. Senior AI test engineers, test leads, and QA managers with CT-AI experience earn $130,000–$180,000 USD. Contract/consulting rates for AI testing specialists range $60–$100+ USD/hour.

Skills validated

Concrete, exam-testable competencies this certification demonstrates.

• ML Model Evaluation Metrics — Accuracy, precision, recall, F1-score, ROC-AUC, confusion matrices, threshold optimization, macro/micro-averaging
• Bias & Fairness Testing — Disparate impact analysis (four-step procedure), sensitive attribute identification, counterfactual generation, equal opportunity testing, calibration analysis
• Data Quality Assessment — Pipeline validation, label correctness verification, class imbalance handling, data representativeness, feature engineering validation
• Adversarial & Robustness Testing — FGSM/PGD attacks, perturbation analysis, boundary value testing, metamorphic testing, out-of-distribution detection
• Neural Network Testing — Activation function behavior, layer-wise testing, neuron coverage, k-neuron coverage, boundary detection coverage
• Generative AI & LLM Testing — Prompt injection detection, jailbreak testing, hallucination assessment, output coherence evaluation, safety guardrail validation
• Concept & Data Drift Detection — Monitoring model performance degradation, statistical drift tests (KL-divergence, Jensen-Shannon), Kolmogorov-Smirnov test
• Test Automation for ML Pipelines — Reproducible test environments, continuous validation frameworks, CI/CD for ML, containerization (Docker)
• A/B Testing & Experimentation — Experiment design, statistical significance testing, sample size calculation, multiple comparison corrections
• Quality Characteristics for AI Systems — Reliability, robustness, fairness verification, explainability (LIME, SHAP), safety and security, privacy-preserving testing
• Requirements Definition — Writing testable acceptance criteria for AI systems, defining edge cases, non-functional requirements (latency, throughput, fairness thresholds)

Career progression with CT-AI

Entry pathway: Typical progression is Foundation Level (CTFL) → Specialist (CT-AI). This is a lateral move in the ISTQB hierarchy; CT-AI does not require CTAL-TA or CTAL-TM but complements them well.

Common career trajectories:

• QA Tester (general) → AI QA Specialist (CT-AI): Expand from testing traditional software into ML-based systems. Estimated experience before CT-AI: 2–4 years in QA.
• Data Analyst / Data Engineer → AI Test Engineer: Data professionals moving into QA discipline; CT-AI validates hybrid skill set. Prerequisites: Python, SQL, statistical literacy.
• Test Lead (CTAL-TM) + CT-AI: Lead AI-focused testing initiatives, design ML testing strategy, mentor junior AI testers. Salary boost: typically 20–30% over non-specialized test managers.
• AI Test Architect / Principal QA Engineer (CT-AI + CTAL-TA): Influence across multiple AI/ML projects, set quality standards, design testing frameworks.

Post-CT-AI certifications to consider:

• ISTQB Certified Tester - Test Automation (CTTA) for automating ML pipeline tests
• ISTQB Certified Tester - Agile Testing (CTAG) for sprint-based AI development
• ISTQB Certified Tester - Testing with Generative AI (CT-GenAI) to specialize further in LLM/generative systems
• Cloud certifications (AWS ML, Azure ML) for MLOps and production deployment context

Industries & roles hiring CT-AI professionals (2026):

• Fintech & banking (fraud detection, credit scoring models)
• Healthcare & biotech (diagnostic AI, drug discovery)
• Autonomous vehicles & robotics (safety-critical ML)
• E-commerce & recommendation engines (personalization testing)
• Government & defense (AI security evaluation)
• Enterprise SaaS (embedded AI features)

Exam preparation strategy

Time investment: Plan 40–60 hours of study, depending on your ML background. Candidates with statistics or ML engineering experience may need less time; testers new to AI/ML should allocate more.

Study phase breakdown:

1. Foundation (10–15 hours): Read the official CT-AI v2.0 syllabus and Itkin's textbook. Focus on ML fundamentals, workflow stages, and testing paradigms unique to AI.
2. Core topics (15–20 hours): Deep-dive into data testing, model testing, neural networks, and generative AI. Work through hands-on exercises.
3. Practice exams (10–15 hours): Complete official sample questions, then move to Udemy practice exams. Aim for 70%+ on practice tests before sitting the real exam.
4. Final review (5–10 hours): Weak areas, flashcards for metrics definitions, and timed mock exams under exam conditions.

Key challenge areas for test-takers:

• Statistical concepts: Expect questions on p-values, confidence intervals, and test significance. Refresh stats fundamentals if rusty.
• Bias vs. variance tradeoff: Understand model behavior across the bias-variance spectrum.
• Metrics selection: Know which metric is appropriate for given scenarios (accuracy vs. F1 vs. AUC-ROC vs. precision-recall curves).
• Generative AI specifics: LLM testing is new in v2.0; allocate extra study time if unfamiliar.

Exam day tips:

• Time limit is tight (60 min / 40 questions = 1.5 min per question). Read carefully but don't overthink.
• Flag questions for review if unsure; come back after completing the rest.
• Passing score is 26/40 (65%); focus on breadth over perfect depth.

Related certifications

• Prerequisite for: None (CT-AI is an endpoint specialist cert; no higher ISTQB cert explicitly requires it, though many advanced professionals stack it with other specialist certs)
• Stacks with: ISTQB Certified Tester - Test Automation (CTTA) (file not yet created) · ISTQB Certified Tester - Agile Testing (CTAG) (file not yet created) · ISTQB Certified Tester - Testing with Generative AI (CT-GenAI) (file not yet created)
• Prerequisite certification: ISTQB Certified Tester Foundation Level (CTFL v4.0) (file not yet created)
• Replaced by: CT-AI v1.0 is retiring; v2.0 is current (v1.0 candidates should transition to v2.0)
• Equivalents at this level: No direct equivalent; comparable AI/ML testing certifications from other vendors are still emerging (as of 2026, ISTQB CT-AI is the most established specialist AI testing credential)
• Vendor overview: ISTQB Overview (file not yet created)

Sources

• ISTQB Certified Tester AI Testing (CT-AI) v2.0
• ISTQB Releases CT-AI Syllabus Version 2.0
• CT-AI Sample Exam Questions v1.3
• ASTQB CT-AI Exam Registration
• ISTQB Exam Pricing FAQ
• Introduction to AI Testing: Guide to ISTQB® CT-AI Certification (Itkin et al.)
• Glassdoor — Software QA Tester Salary
• PayScale — QA Tester Salary (US)
• PayScale — Software Tester Salary (South Africa)
• Indeed — Software Quality Assurance Tester Salaries
• Udemy — ISTQB AI Testing Masterclass
• ProcessExam — CT-AI Sample Questions

Last verified: 2026-05-01
Parent ecosystem: ISTQB Testing Ecosystem (file not yet created)
Parent domain: Quality Assurance (file not yet created)
Vendor overview: ISTQB Overview (file not yet created)