Updated Sep-2026 100% Cover Real CT-AI Exam Questions Make Sure You 100% Pass [Q97-Q116]

5/5 - (1 vote)

Updated Sep-2026 100% Cover Real CT-AI Exam Questions Make Sure You 100% Pass

CT-AI dumps Accurate Questions and Answers with Free and Fast Updates

ISTQB CT-AI Exam Syllabus Topics:

Topic	Details
Topic 1	Introduction to AI: This exam section covers topics such as the AI effect and how it influences the definition of AI. It covers how to distinguish between narrow AI, general AI, and super AI; moreover, the topics covered include describing how standards apply to AI-based systems.
Topic 2	Testing AI-Specific Quality Characteristics: In this section, the topics covered are about the challenges in testing created by the self-learning of AI-based systems.
Topic 3	ML Functional Performance Metrics: In this section, the topics covered include how to calculate the ML functional performance metrics from a given set of confusion matrices.
Topic 4	Test Environments for AI-Based Systems: This section is about factors that differentiate the test environments for AI-based
Topic 5	Testing AI-Based Systems Overview: In this section, focus is given to how system specifications for AI-based systems can create challenges in testing and explain automation bias and how this affects testing.
Topic 6	Machine Learning ML: This section includes the classification and regression as part of supervised learning, explaining the factors involved in the selection of ML algorithms, and demonstrating underfitting and overfitting.
Topic 7	Quality Characteristics for AI-Based Systems: This section covers topics covered how to explain the importance of flexibility and adaptability as characteristics of AI-based systems and describes the vitality of managing evolution for AI-based systems. It also covers how to recall the characteristics that make it difficult to use AI-based systems in safety-related applications.
Topic 8	Methods and Techniques for the Testing of AI-Based Systems: In this section, the focus is on explaining how the testing of ML systems can help prevent adversarial attacks and data poisoning.

QUESTION 97
Which characteristic of AI-based systems makes it difficult to ensure they are safe (e.g., not harming humans)?
Choose ONE option (1 out of 4)

Determinism

Complexity

Interpretability

Robustness

QUESTION 98
Which of the following is one of the reasons for data mislabelling?

Lack of domain knowledge

Expert knowledge

Interoperability error

Small datasets

QUESTION 99
Which ONE of the following describes a situation of back-to-back testing the LEAST?
SELECT ONE OPTION

Comparison of the results of a current neural network model ML model implemented in platform A (for example Pytorch) with a similar neural network model ML model implemented in platform B (for example Tensorflow), for the same data.

Comparison of the results of a home-grown neural network model ML model with results in a neural network model implemented in a standard implementation (for example Pytorch) for same data

Comparison of the results of a neural network ML model with a current decision tree ML model for the same data.

Comparison of the results of the current neural network ML model on the current data set with a slightly modified data set.

Back-to-back testing is a method where the same set of tests are run on multiple implementations of the system to compare their outputs. This type of testing is typically used to ensure consistency and correctness by comparing the outputs of different implementations under identical conditions. Let’s analyze the options given:
A . Comparison of the results of a current neural network model ML model implemented in platform A (for example Pytorch) with a similar neural network model ML model implemented in platform B (for example Tensorflow), for the same data.
This option describes a scenario where two different implementations of the same type of model are being compared using the same dataset. This is a typical back-to-back testing situation.
B . Comparison of the results of a home-grown neural network model ML model with results in a neural network model implemented in a standard implementation (for example Pytorch) for the same data.
This option involves comparing a custom implementation with a standard implementation, which is also a typical back-to-back testing scenario to validate the custom model against a known benchmark.
C . Comparison of the results of a neural network ML model with a current decision tree ML model for the same data.
This option involves comparing two different types of models (a neural network and a decision tree). This is not a typical scenario for back-to-back testing because the models are inherently different and would not be expected to produce identical results even on the same data.
D . Comparison of the results of the current neural network ML model on the current data set with a slightly modified data set.
This option involves comparing the outputs of the same model on slightly different datasets. This could be seen as a form of robustness testing or sensitivity analysis, but not typical back-to-back testing as it doesn’t involve comparing multiple implementations.
Based on this analysis, option C is the one that describes a situation of back-to-back testing the least because it compares two fundamentally different models, which is not the intent of back-to-back testing.

QUESTION 100
The following confusion matrix represents the functional performance of a classifier.

Which ONE of the following is the correct calculation for the accuracy of the classifier?

60 / (60 + 20) * 100% = 75%

60 / (60 + 9) * 100% = 87%

2 * (60 / (60 + 20) * 60 / (60 + 9)) / (60 / (60 + 20) + 60 / (60 + 9)) * 100% = 80%

(60 + 11) / (60 + 11 + 20 + 9) * 100% = 71%

QUESTION 101
The stakeholders of a machine learning model have confirmed that they understand the objective and purpose of the model, and ensured that the proposed model aligns with their business priorities. They have also selected a framework and a machine learning model that they will be using. What should be the next step to progress along the machine learning workflow?

Tune the machine learning algorithm based on objectives and business priorities

Prepare and pre-process the data that will be used to train and test the model

Agree on defined acceptance criteria for the machine learning model

Evaluate the selection of the framework and the model

QUESTION 102
A test engineer is planning testing for a wearable medical device using AI. The medical device will detect possible heart issues in patients and dispatch emergency services automatically. It is not expected that many patients will have heart issues, and this is reflected in the available data.
In this case, it is decided it is more important that emergency services are not sent un- necessarily, than it is to detect actual heart problems.
Which ONE of the following metrics should the test engineer choose to ensure that the emergency services are sent only when needed?

Accuracy

MSE

R-Squared

Precision

QUESTION 103
Which ONE of the following statements correctly describes the importance of flexibility for Al systems?

Al systems are inherently flexible.

Al systems require changing of operational environments; therefore, flexibility is required.

Flexible Al systems allow for easier modification of the system as a whole.

Self-learning systems are expected to deal with new situations without explicitly having to program for it.

QUESTION 104
A beer company is trying to understand how much recognition its logo has in the market. It plans to do that by monitoring images on various social media platforms using a pre-trained neural network for logo detection.
This particular model has been trained by looking for words, as well as matching colors on social media images. The company logo has a big word across the middle with a bold blue and magenta border.
Which associated risk is most likely to occur when using this pre-trained model?

There is no risk, as the model has already been trained

Insufficient function; the model was not trained to check for colors or words

Improper data preparation

Inherited bias: the model could have inherited unknown defects

QUESTION 105
A robotic AI-based system is being built by a logistics company to operate within its unmanned warehouses. The warehouses can all be very different and new ones are being added each year.
It is expected that the system will not require retraining for each warehouse, and will be able to learn the location of different items and move them to specified locations on request.
Which ONE of the following attributes should be MOST carefully considered when specifying the objectives and acceptance criteria for the system?

Inappropriate bias

Adaptability

Ethics

Explainability

QUESTION 106
Which ONE of the following tests is MOST likely to describe a useful test to help detect different kinds of biases in ML pipeline?

Testing the distribution shift in the training data for inappropriate bias.

Test the model during model evaluation for data bias.

Testing the data pipeline for any sources for algorithmic bias.

Check the input test data for potential sample bias.

QUESTION 107
“BioSearch” is creating an Al model used for predicting cancer occurrence via examining X-Ray images. The accuracy of the model in isolation has been found to be good. However, the users of the model started complaining of the poor quality of results, especially inability to detect real cancer cases, when put to practice in the diagnosis lab, leading to stopping of the usage of the model.
A testing expert was called in to find the deficiencies in the test planning which led to the above scenario.
Which ONE of the following options would you expect to MOST likely be the reason to be discovered by the test expert?
SELECT ONE OPTION

A lack of similarity between the training and testing data.

The input data has not been tested for quality prior to use for testing.

A lack of focus on choosing the right functional-performance metrics.

A lack of focus on non-functional requirements testing.

QUESTION 108
Which of the following is correct regarding the layers of a deep neural network?

There is only an input and output layer

There is at least one internal hidden layer

There must be a minimum of five total layers to be considered deep

The output layer is not connected with the other layers to maintain integrity

QUESTION 109
“BioSearch” is creating an Al model used for predicting cancer occurrence via examining X-Ray images. The accuracy of the model in isolation has been found to be good. However, the users of the model started complaining of the poor quality of results, especially inability to detect real cancer cases, when put to practice in the diagnosis lab, leading to stopping of the usage of the model. A testing expert was called in to find the deficiencies in the test planning which led to the above scenario.
Which ONE of the following options would you expect to MOST likely be the reason to be discovered by the test expert?

A lack of similarity between the training and testing data.

The input data has not been tested for quality prior to use for testing.

A lack of focus on choosing the right functional-performance metrics.

A lack of focus on non-functional requirements testing.

QUESTION 110
You have access to the training data that was used to train an AI-based system. You can review this information and use it as a guideline when creating your tests. What type of characteristic is this?

Autonomy

Explorability

Transparency

Accessibility

QUESTION 111
When verifying that an autonomous AI-based system is acting appropriately, which of the following are MOST important to include?

Test cases to verify that the system automatically confirms the correct classification of training data

Test cases to detect the system appropriately automating its data input

Test cases to detect the system prompting for unnecessary human intervention

Test cases to verify that the system automatically suppresses invalid output data

QUESTION 112
Which AI-specific test objective and acceptance criterion should be selected MOST LIKELY for testing GPT_Legal?

Test objective: Evidence of functional safety Acceptance criterion: The system recognizes failures in the transmission of information and data with the DPMA system and the evaluation system by means of self-tests.

Test objective: Evidence of evolution Acceptance criterion: The quality of the research results does not deteriorate with further training.

Test objective: Evidence of compatibility Acceptance criterion: The system can exchange information with the DPMA system and the evaluation system.

Test objective: Evidence that the data is free from inappropriate bias Acceptance criterion: The DPMA’s analysis data is statistically compared to data from other sources.

QUESTION 113
A wildlife conservation group would like to use a neural network to classify images of different animals. The algorithm is going to be used on a social media platform to automatically pick out pictures of the chosen animal of the month. This month’s animal is set to be a wolf. The test team has already observed that the algorithm could classify a picture of a dog as being a wolf because of the similar characteristics between dogs and wolves. To handle such instances, the team is planning to train the model with additional images of wolves and dogs so that the model is able to better differentiate between the two. What test method should you use to verify that the model has improved after the additional training?

Metamorphic testing because the application domain is not clearly understood at this point.

Adversarial testing to verify that no incorrect images have been used in the training.

Pairwise testing using combinatorics to look at a long list of photo parameters.

Back-to-back testing using the version of the model before training and the new version of the model after being trained with additional images.

QUESTION 114
Which of the following options BEST describes an example of reinforcement learning?

A house cleaning robot adjusts the path it takes on a daily basis based on the most efficient route and the amount of dirt it is able to remove.

A financial system that chooses which stocks to buy regularly optimizes based on new changes in stock price to improve its performance.

An image recognition system regularly searches the internet to find new examples to add to its training data.

A health warning system uses data from thousands of people to predict who is most likely to suffer from a particular health condition based on their behaviour.

QUESTION 115
“AllerEgo” is a product that uses sell-learning to predict the behavior of a pilot under combat situation for a variety of terrains and enemy aircraft formations. Post training the model was exposed to the real- world data and the model was found to be behaving poorly. A lot of data quality tests had been performed on the data to bring it into a shape fit for training and testing.
Which ONE of the following options is least likely to describes the possible reason for the fall in the performance, especially when considering the self-learning nature of the Al system?
SELECT ONE OPTION

The difficulty of defining criteria for improvement before the model can be accepted.

The fast pace of change did not allow sufficient time for testing.

The unknown nature and insufficient specification of the operating environment might have caused the poor performance.

There was an algorithmic bias in the Al system.

QUESTION 116
Which of the following is a dataset issue that can be resolved using pre-processing?

Insufficient data

Invalid data

Wanted outliers

Numbers stored as strings