Are these real CT-GenAI exam questions?

No — and you should be wary of any site that says its questions are. Real exam items are confidential. These questions are written from scratch against the published CT-GenAI exam guide to match its style, difficulty and domain weighting.

Should I use CT-GenAI dumps?

No. A dump is a collection of leaked or memorised real exam items, and using one breaks the ISTQB certification agreement — the vendor can revoke your certification and ban you from future exams. Dumps also tend to fail on their own terms: their answer keys are unverified, they drift out of date as the exam changes, and memorising answers leaves you stuck the moment a question is reworded. Original practice questions with an explanation on every option teach the reasoning the CT-GenAI actually tests, which is what carries over to questions you have never seen.

Do the CT-GenAI questions come with explanations?

Every option carries its own explanation — not just the correct one. You find out why the answer you nearly picked was wrong, which is where most of the learning actually happens. Each question also links back to the revision lesson that teaches the concept.

How many practice questions should I do before the CT-GenAI exam?

Volume matters less than what you do with the misses. A useful signal is consistency rather than a total: when you can hold a comfortable margin above the 30 pass mark on questions you have never seen before, across every domain rather than just your strong ones, you are ready.

Can I retry the questions I got wrong?

Yes. Wrong answers and flagged questions go into a review pool you can drill on their own, so you can grind exactly the material that is costing you marks instead of re-answering the ones you already know.

CT-GenAI practice

Free CT-GenAI practice questions

Drill exam-realistic ISTQB Certified Tester Testing with Generative AI (CT-GenAI) questions by domain, with an explanation on every option — not just the right one. 5 fully worked examples are further down this page, answers included.

Question bank: 416
across 5 domains
Free, no account: 5/day
a free account allows 10 a day
Real exam: 40 Qs
60 min · pass 30 / 46 (65%)
Explanations: Every option
right and wrong

Build a practice session

5 free questions left today.

Domains

How many?

Mode

Ready when you are

10 fresh questions drawn across all domains, in Learn mode.

Review what you missed

Every question you answer incorrectly, and every question you flag while practising, is saved here automatically. Finish a session and you can come back to re-drill just those.

5 sample CT-GenAI questions, fully explained

Real questions from the CT-GenAI bank, with the answer key and the reasoning behind every option. Read them, then go back up the page and try the rest.

Question 1Introduction to Generative AI for Software Testing

A test team uses a tool in which domain experts write explicit if-then statements that decide whether a test result should be reported as a defect. Which type of AI BEST describes this tool?

Choose one.

a
Classical machine learning
Classical machine learning learns its decision logic from prepared data and selected features during training; nobody writes the rules by hand.
b
Deep learning
Deep learning trains neural networks that learn features from large datasets; the decision logic is learned, not written as explicit statements.
c
Symbolic AI Correct
Symbolic AI represents knowledge with symbols and logical rules written by people, which is exactly what expert-authored if-then statements are.
d
Generative AI
Generative AI uses deep learning to create new content such as text or code; it does not decide outcomes from hand-written rules.

The concept

Symbolic AI mimics human decision-making with knowledge encoded as symbols and logical rules.

Why that’s the answer

The deciding detail is who supplies the logic. In symbolic AI, experts encode their knowledge as explicit rules that the system applies. Classical machine learning and deep learning both learn their logic from data during training, and generative AI creates new content from patterns learned in pre-training.

How to reason it out

Ask where the tool's decision logic comes from.
Logic written by experts as rules points to symbolic AI.
Logic learned from data points to machine learning or deep learning.
Creating new content points to generative AI.

Exam tip: Hand-written rules and symbols = symbolic AI; learned from data = machine learning.

CT-GenAI: Generative AI and LLM Fundamentals for Software Testing — the lesson that teaches this.

Question 2Prompt Engineering for Effective Software Testing

A tester begins a prompt with: "Act as a senior test manager reviewing a draft test plan." Which prompt component does this sentence MAINLY provide?

Choose one.

a
Context
Context is background about the test object and the functionality under test. The sentence describes who the model should be, not the system being tested.
b
Instruction
Although it starts with a verb, the sentence does not state the task to perform. The task (for example, list the plan's gaps) would be the instruction.
c
Constraints
Constraints restrict how the task is carried out on the input data. Nothing in the sentence limits or restricts the response.
d
Role Correct
The sentence sets the persona and perspective the model should adopt, which is the role component. It shapes the responsibilities and tone the LLM takes on.

The concept

The role component tells the LLM which perspective or persona to adopt, such as tester, test manager or test automation engineer.

Why that’s the answer

In the six-component prompt structure, the role defines the persona the model takes when it responds, which helps it pick suitable responsibilities and tone. "Act as a senior test manager" is a classic role statement. The rest of the sentence only tells the persona what it is doing in general; the specific task, the plan itself and any restrictions would be separate instruction, input data and constraint components.

How to reason it out

Ask what the sentence tells the model: who to be, what to know, what to do, or what not to do.
"Act as …" defines a persona, which is the role.
Check that no task, background on the test object, or restriction is actually stated.

Exam tip: "Act as / You are a …" = role: the perspective the LLM takes.

CT-GenAI: Prompt Structure, Prompting Techniques and System vs User Prompts — the lesson that teaches this.

Question 3Managing Risks of Generative AI in Software Testing

Which description BEST matches a hallucination in LLM output produced for a software test task?

Choose one.

a
Output that misapplies cause-and-effect or conditional logic to reach a conclusion
Misinterpreting logical structures such as conditions or cause and effect is a reasoning error, not a hallucination.
b
Output that systematically favours certain assumptions inherited from training data
A systematic tilt that comes from the training data is a bias. A hallucination is incorrect or irrelevant content, not a consistent preference.
c
Output that differs between two runs of the same prompt with identical input data
Variation between runs is the non-deterministic behaviour of LLMs. It is not itself a defect in the content of any one response.
d
Output that appears factually incorrect or irrelevant to the given task Correct
This is the syllabus definition of a hallucination: content that looks plausible but is wrong or does not belong to the task, such as a test case for a feature that does not exist.

The concept

A hallucination is LLM output that appears factually incorrect or irrelevant to the task it was given.

Why that’s the answer

The syllabus defines hallucinations as output that appears factually incorrect or irrelevant to a given task. In testing they show up as fictitious or irrelevant test cases, incorrect or non-functioning test scripts, or tests that verify acceptance criteria that do not exist. The other descriptions name the two sibling defect types (reasoning errors and biases) and the separate property of non-deterministic output.

How to reason it out

Separate the three defect types: hallucination, reasoning error, bias.
Hallucination = incorrect or irrelevant content.
Faulty logic = reasoning error; skew from training data = bias; run-to-run variation = non-determinism.

Exam tip: Invented or irrelevant content = hallucination.

CT-GenAI: Hallucinations, Reasoning Errors and Bias in LLM Testing — the lesson that teaches this.

Question 4LLM-Powered Test Infrastructure for Software Testing

Which statement BEST describes an LLM-powered test infrastructure, as opposed to a traditional AI chatbot?

Choose one.

a
A system that lets testers hold open-ended conversations with an LLM to clarify test concepts
Conversational interaction is the primary focus of a traditional AI chatbot, which is exactly what the test infrastructure is contrasted with.
b
A system that integrates an LLM into the test process to support automation, reasoning and decisions Correct
This is how the syllabus defines LLM-powered test infrastructure: the LLM is built into the test process to enhance automation, reasoning and decision-making on test tasks such as analysing requirements and generating test cases.
c
A system that answers test questions from scripted responses chosen by predefined matching rules
Scripted, rule-based responses describe a rule-based chatbot. An LLM-powered infrastructure generates insights dynamically from context.
d
A system that sends the tester's typed prompt unchanged to an LLM and then shows its raw response
The typical architecture has a back end that prepares prompts and post-processes the raw output; passing prompts and responses through untouched leaves out the components that make it test infrastructure.

The concept

An LLM-powered test infrastructure integrates an LLM into the test process to enhance automation, reasoning and decision-making, rather than focusing on conversation.

Why that’s the answer

The syllabus contrasts LLM-powered test infrastructure with a traditional AI chatbot: the chatbot primarily focuses on conversational interaction, while the test infrastructure is designed to support testing by processing test-related queries, analysing requirements, generating test cases and evaluating outputs. Scripted rule-based answers and an unprocessed pass-through to the LLM both miss what the architecture adds.

How to reason it out

Identify what the system is built for: conversation, or work inside the test process.
Test infrastructure processes test queries, analyses requirements, generates and evaluates test artefacts.
Reject descriptions of chatbots, rule-based bots or a bare pass-through to the model.

Exam tip: Chatbot = conversation; LLM-powered test infrastructure = LLM built into the test process.

CT-GenAI: LLM-Powered Test Infrastructure, RAG and LLM Agents — the lesson that teaches this.

Question 5Deploying and Integrating Generative AI in Test Organizations

Which statement BEST defines shadow AI in a test organization?

Choose one.

a
GenAI features running unnoticed inside test tools the organization has already approved
Features inside an approved tool fall under that tool's approval. Shadow AI is about tools nobody approved or oversees.
b
GenAI tools or systems used in the organization without formal approval or oversight Correct
This is the syllabus glossary definition: the defining feature is the lack of formal approval or oversight, not where the tool runs or what it produces.
c
GenAI output placed into testware with no marking to show it was generated by a model
Unmarked output breaks a transparency obligation of the GenAI process guidelines; it is not what shadow AI means.
d
GenAI models that were trained on data the organization has not yet reviewed or classified
Unreviewed training data is a data-quality or governance concern, not the definition of shadow AI.

The concept

Shadow AI is the use of GenAI tools or systems within an organization without formal approval or oversight.

Why that’s the answer

The syllabus defines shadow AI by its governance gap: the tool is used without formal approval or oversight. Unmarked generated content and unreviewed training data are real issues, but they are covered by transparency rules and data quality, and a feature inside an approved tool is not unapproved.

How to reason it out

Look for the governance gap in each option.
Shadow AI means no formal approval and no oversight.
Reject options about transparency or data quality, which are separate concerns.

Exam tip: Shadow AI = GenAI used without formal approval or oversight.

CT-GenAI: GenAI Adoption Roadmap, Shadow AI and LLM Selection — the lesson that teaches this.

The CT-GenAI question bank, by domain

The bank is built to the exam's own weighting, so the practice you get reflects the marks that are actually on offer — not whichever domain was easiest to write questions for.

Published CT-GenAI practice questions per exam domain
Domain	Exam weight	Topics	Questions
Introduction to Generative AI for Software Testing	17%	2	74
Prompt Engineering for Effective Software Testing	28%	3	114
Managing Risks of Generative AI in Software Testing	25%	3	104
LLM-Powered Test Infrastructure for Software Testing	13%	2	54
Deploying and Integrating Generative AI in Test Organizations	17%	2	70
Total	100%	12	416

Practise CT-GenAI one domain at a time

How CT-GenAI questions are worded

Most CT-GenAI questions are not asking whether you can recall a definition. They describe a situation and ask which option satisfies it — so the skill being tested is reading the requirement precisely and eliminating options that fail it.

Single-response vs multiple-response

A single-response question has exactly one right answer. A multiple-response question tells you how many to pick ("Choose TWO") and there is no partial credit — getting one of the two right scores nothing. Read that instruction before you read the options.

Read the last line first

The final sentence is the actual question; everything before it is scenario. Read it first, then read the scenario knowing what you are looking for. It stops you from building an answer in your head that the question never asked for.

Hunt for the qualifier

Most scenarios turn on one word — MOST cost-effective, LEAST operational overhead, with the LEAST latency, without changing application code. Two options are frequently both technically correct, and the qualifier is the only thing separating them.

Eliminate, then choose

Distractors are almost always real ISTQB services doing a real job — just not this job. Rule out the ones that break a stated constraint before you compare what's left. On a question you truly don't know, eliminating two options turns a guess into a coin flip.

Practise CT-GenAI from your browser

The SaveMyCert Chrome extension opens a daily drill of 5 CT-GenAI questions in the side panel, with the full explanation behind every answer. No account needed.

Get the Chrome extension

Beyond CT-GenAI practice

CT-GenAI practice questions: your questions

Yes. You can answer 5 questions a day with no account at all, and a free account raises that to 10 a day across the full 416-question CT-GenAI bank, with an explanation on every option. Pro removes the daily cap entirely.

Source

Exam structure, domain weights and scoring on this page come from the official CT-GenAI exam guide.

ISTQB Certified Tester Testing with Generative AI (CT-GenAI) — exam structure and syllabus — ISTQB