Checked 29 September 2026. Built on CT-AI Syllabus v2.0 (minutes per chapter, hands-on objectives) and the Exam Structures & Rules Tables v1.19 (points). The plan itself is our method; the booking lead time is ITB's current procedure.
ISTQB sets a minimum training time for CT-AI v2.0 of 1,170 minutes — 19.5 hours, down from 25.1 hours for v1.0 (ISTQB describes it as "reduced to 3 days of training"). That figure is for accredited courses, but it is also a sensible self-study budget, and the syllabus tells you exactly how it is split across the seven chapters. The catch is that teaching time and exam points do not line up: chapter 3 has the longest allocation, 32% of the minutes, and also holds five of the eight hands-on exercises, but carries 18% of the points, and hands-on exercises are not examined. This guide turns the official numbers into a four-week plan and shows what changes when you study in points order.
CT-AI Syllabus v2.0 (0.6, 0.10); Sample Exam A v2.2; CT-AI v2.0 FAQ.
Where the 19.5 hours go, and what they earn
| Chapter | Minutes (syllabus) | Share of time (ours) | Points | Share of points (ours) | Points per hour (ours) |
|---|---|---|---|---|---|
| 1. Introduction to AI | 120 | 10.3% | 6 | 13.6% | 3.0 |
| 2. Quality characteristics for AI-based systems | 45 | 3.8% | 3 | 6.8% | 4.0 |
| 3. Machine learning | 375 | 32.1% | 8 | 18.2% | 1.3 |
| 4. Testing AI-based systems | 195 | 16.7% | 8 | 18.2% | 2.5 |
| 5. Input data testing for ML systems | 180 | 15.4% | 7 | 15.9% | 2.3 |
| 6. Model testing for ML systems | 225 | 19.2% | 10 | 22.7% | 2.7 |
| 7. ML development testing | 30 | 2.6% | 2 | 4.5% | 4.0 |
| Total | 1,170 | 100% | 44 | 100% | 2.3 |
Minutes from CT-AI Syllabus v2.0 section 0.10; points from ISTQB Exam Structures & Rules Tables v1.19; shares and rates are our arithmetic, rounded to one decimal place, so a column may not add to exactly 100%.
Read the last column as a priority signal, not a rule. Chapters 2 and 7 are cheap points (4 per training hour) — short, vocabulary-heavy, easy to leave until the end and then forget. Chapter 3 is the expensive one, but it also contains the confusion-matrix K3 question, so don't skip it; skip its exercises if you are short of time.
Study order changes how fast you reach the pass mark
Our model: a chapter counts only when all its syllabus minutes are done. It is not a prediction of your score.
| Hours studied | Points covered, syllabus order (ours) | Points covered, points-per-hour order (ours) |
|---|---|---|
| 4 | 9 | 11 |
| 8 | 9 | 21 |
| 12 | 17 | 29 |
| 16 | 32 | 36 |
| 19.5 | 44 | 44 |
| Hours until chapters worth 29 points are covered | 15.25 h | 10.25 h |
Our arithmetic on the syllabus minutes and exam points. Points-per-hour order is chapters 2, 7, 1, 6, 4, 5, 3. Covering 29 points of material is not the same as scoring 29; the point is that front-loading cheap chapters protects you if your preparation gets cut short.
A four-week plan
| Week | Chapters | Syllabus minutes | Points | Also do |
|---|---|---|---|---|
| 1 | 2 Quality characteristics, 7 ML development testing, 1 Introduction to AI | 195 | 11 | Learn every keyword list at the start of each chapter: the syllabus (0.5) says names and definitions must be remembered (K1), and ISTQB's exam rules (4.1.4) examine them for understanding (K2) |
| 2 | 6 Model testing | 225 | 10 | Metamorphic testing (K3) until you can write three MRs for any scenario |
| 3 | 4 Testing AI-based systems, 5 Input data testing | 375 | 15 | Red-teaming steps and the ten dataset constraint types (both K3) |
| 4 | 3 Machine learning, then revision | 375 | 8 | Confusion-matrix metrics (K3); Sample Exam A v2.2 under 60 minutes; re-read weak chapters |
| Total | All seven | 1,170 | 44 | Plus 60 minutes for the sample exam |
Our plan, in points-per-hour order, using ISTQB's minutes and points. About 5 hours a week. If your exam is booked through ITB, remember the slot has to be at least 3 days after payment, Monday to Friday.
The eight hands-on exercises: useful, but not examined
The syllabus is explicit: "All syllabus sections are examinable except for the Introduction, Hands-On Objectives, References, and Appendices." Accredited courses must run the exercises. If you are self-studying for the exam only, read what each one is meant to show; if you want the skill, the exercises are where most of the learning is.
| Hands-on objective | Exercise | Level | Supports |
|---|---|---|---|
| HO-3.1.3 | Create an ML model | H2 | Chapter 3 |
| HO-3.2.2 | Data preparation | H2 | Chapter 3 |
| HO-3.3.2 | Evaluate a model using selected functional performance metrics | H2 | AI-3.3.1 (K3) |
| HO-3.3.3 | Show the impact of different model and dataset combinations | H2 | Chapter 3 |
| HO-3.4.2 | Implement a perceptron | H1 | Chapter 3 neural networks |
| HO-4.2.3 | Exploratory testing of an LLM generating boundary value analysis test cases | H2 | Chapter 4 GenAI testing |
| HO-5.1.7 | Input data testing | H2 | AI-5.1.5 (K3) |
| HO-6.1.6 | Apply metamorphic testing | H2 | AI-6.1.5 (K3) |
CT-AI Syllabus v2.0, sections 0.6 and the hands-on sections. Five of the eight sit in chapter 3, which also has the longest allocation, 375 of the 1,170 minutes. If you do only three, do the ones behind the K3 objectives: HO-3.3.2, HO-5.1.7 and HO-6.1.6.
Your plan
Your CT-AI v2.0 plan: hours per chapter and the date to pay by
Enter the date you want to sit and your study hours per week. The planner splits your hours half by exam points and half by ISTQB's training minutes, lists the chapters in points-per-hour order (the order this guide recommends), and works out the latest date to pay the Indian Testing Board under its 3-day minimum if you book with them.
Points from ISTQB's Exam Structures & Rules Tables v1.19; minutes from Syllabus v2.0 section 0.10. The 50/50 blend and the pay-by date are our method. ITB asks for an exam date at least 3 days after payment, Monday to Friday, and confirms the slot within 48 hours; leave more room than the minimum.
| Chapter | Planner weight (ours) | Hours out of 19.5 (ours) |
|---|---|---|
| 1. Introduction to AI | 11.9% | 2.3 |
| 2. Quality characteristics for AI-based systems | 5.3% | 1.0 |
| 3. Machine learning | 25.1% | 4.9 |
| 4. Testing AI-based systems | 17.4% | 3.4 |
| 5. Input data testing for ML systems | 15.6% | 3.1 |
| 6. Model testing for ML systems | 21.0% | 4.1 |
| 7. ML development testing | 3.6% | 0.7 |
The planner's weights: half by the chapter's share of the 44 points, half by its share of the 1,170 training minutes. The weights set the hours; the list order is points per training hour, as in the four-week plan. Our method.
Free official resources
| Resource | Where | Use it for | Downloads shown (29 Sep 2026) |
|---|---|---|---|
| CT-AI Syllabus v2.0 (93 pages, GA 17 Apr 2026) | ISTQB CT-AI page | The core examinable text: LOs, minutes per chapter, keywords (their definitions come from the ISTQB Glossary or the syllabus's chapter 9) | 22,174 |
| Sample Exam A v2.2, questions | ISTQB CT-AI page | 40 exam questions plus 6 additional ones | 9,127 |
| Sample Exam A v2.2, answers (released 20 Jul 2026) | ISTQB CT-AI page | Answer key with the LO and K-level for each question and explanations | 7,124 |
| Accreditation Guidelines | ISTQB CT-AI page | For training providers; the linked file is still v1.0 (2021) | 10,146 |
| Exam Structures & Rules Tables v1.19 | ISTQB downloads | Questions and points per chapter | — |
ISTQB CT-AI page and download counters as we read them on 29 September 2026; the counters rise continuously, so treat them as a snapshot. Make sure a sample exam says v2.x: the sample was revised twice after release (v2.1 on 20 May 2026 updated question 10; v2.2 on 20 Jul 2026 replaced question 19). ISTQB warns sample questions "may both be more difficult or less difficult" than the real exam.
Last-week checklist
- Take Sample Exam A v2.2 once, timed at 60 minutes. Aim for well above 29 of 44, not just at it.
- For every wrong answer, find the LO in the answer key and re-read that syllabus section, not just the explanation.
- Practise reading a confusion matrix drawn both ways round.
- Write the red-teaming steps, the ten constraint types and the three MR properties from memory once a day.
- Book with a margin: ITB wants the exam at least 3 days after payment and confirms within 48 hours.
- Check your webcam and connection against ITB's Candidate Criteria Checklist before you pay, not on exam day.
Our checklist, built on ISTQB's sample-exam notes and ITB's remote-proctored procedure.
Practise it, don't just read it
Ten ProSyllabus quizzes of 10 questions each map to the seven v2.0 chapters (two each for chapters 1, 3, 4 and 6; the second chapter 6 quiz also covers chapter 7), so you can test each chapter the week you finish it.
The ProSyllabus CT-AI library has 10 quizzes of 10 questions written on the v2.0 syllabus and covering all seven chapters: Introduction to AI (two quizzes), quality characteristics, machine learning (two: forms and workflow; metrics, neural networks and coverage), testing AI-based systems (two: test oracles and test levels; generative AI and red teaming), input data testing, and model testing plus ML development testing (two). They are our own questions, not ISTQB exam items, and they are single-answer only, while ISTQB's official Sample Exam A v2.2 also includes two select-two questions, so practise that format there.
How many hours does it take to prepare for CT-AI v2.0?
ISTQB sets a minimum of 1,170 training minutes, 19.5 hours, for accredited courses. Self-study time depends on your background; it is a sensible minimum budget.
Can I self-study for CT-AI without a course?
ISTQB says it highly recommends accredited training, but it is not required to sit the exam. The syllabus and Sample Exam A v2.2 are free on ISTQB's site.
Are the CT-AI hands-on exercises in the exam?
No. The syllabus says all sections are examinable except the introduction, hands-on objectives, references and appendices.
Which chapter should I study first?
Our suggestion is points per hour: chapters 2, 7 and 1 first, then 6, 4, 5 and finally 3, which takes the most time for its 8 points.
Which chapter takes the most study time?
Chapter 3, Machine Learning: 375 of the 1,170 minutes. It also holds five of the eight hands-on exercises.
Is there an official CT-AI v2.0 mock test?
Yes. ISTQB's Sample Exam A v2.2 has 40 questions plus 6 additional ones, with an answer key released on 20 July 2026.
Do I need Python or ML experience?
No prior AI knowledge is required, per ISTQB's FAQ. The hands-on exercises use ML tools, but they are not examined; some familiarity with testing or development is recommended.
How early should I book the exam in India?
ITB asks for an exam date at least 3 days after payment, Monday to Friday, and confirms within 48 hours. Leave more margin than that.
Sources
- official — ISTQB — CT-AI Syllabus v2.0 (GA 17 Apr 2026, PDF)
- official — ISTQB — Exam Structures & Rules Tables v1.19 (19 Aug 2026, PDF)
- official — ISTQB — Certified Tester AI Testing (CT-AI) certification page (v2.0)
- official — ISTQB — Help: CT-AI Version 2.0 (FAQ)
- official — ISTQB — CT-AI Sample Exam set A, Questions v2.2 (PDF)
- official — ISTQB — CT-AI Sample Exam set A, Answers v2.2 (PDF, released 20 Jul 2026)
- official — ISTQB — CT-AI Accreditation Guidelines v1.0 (2021, PDF linked from the v2.0 page)
- official — ISTQB — News: ISTQB Releases Certified Tester AI Testing (CT-AI) Syllabus Version 2.0 (21 Apr 2026)
- official — Indian Testing Board — Specialist: Remote proctored exam (take exam from home), procedure and fee table
official = a document published by the conducting body. reported = a news or coaching site we could not check against an original. Where sources disagree this page says so rather than picking one.
