A tech company is developing a facial recognition system. The system is trained on a large dataset of customer images. The privacy team is concerned about the risk of data leakage through the model, where an attacker could extract training images. The company must comply with data protection regulations and wants to minimize the risk of this attack. Which approach is most effective?

Use pseudonymization to replace image file names with random IDs.

Use differential privacy to add noise to the training process.

Use federated learning to train the model on user devices.

Use homomorphic encryption to encrypt the training images.

What is the primary purpose of consent tagging in a data processing system?

To encrypt personal data at rest and in transit

To provide a user interface for submitting data subject access requests

To associate a user's consent choices with specific data processing activities

To anonymize personal data so it can no longer identify an individual

What is a common challenge when implementing consent tagging across multiple systems?

Providing a public API for third-party developers

Ensuring that consent tags are consistently applied and synchronized across all systems that process the data

Ensuring that data is stored in a single centralized database

Generating cryptographic keys for each user

Which of the following is an example of a Privacy-Enhancing Technology?

Differential privacy, which adds noise to query results

A backup system that copies data to a remote server

An intrusion detection system that monitors for attacks

A firewall that blocks unauthorized network access

What is the primary purpose of differential privacy in AI models?

To encrypt the model's parameters

To add noise to data or queries to prevent re-identification of individuals

To make the model's predictions more accurate

Certified Data Privacy Solutions Engineer

Domain 4Objective 3

Privacy-Enhancing Technologies CDPSE Practice Questions (Page 4)

Part of the Privacy Engineering domain, which accounts for 39% of the CDPSE exam.

31questions here

7free pages

10concepts

39%of the exam

Questions 16–20

16expert · hard
A tech company is developing a facial recognition system. The system is trained on a large dataset of customer images. The privacy team is concerned about the risk of data leakage through the model, where an attacker could extract training images. The company must comply with data protection regulations and wants to minimize the risk of this attack. Which approach is most effective?
Select an answer first
17foundation · easy
What is the primary purpose of consent tagging in a data processing system?
Select an answer first
18foundation · easy
What is a common challenge when implementing consent tagging across multiple systems?
Select an answer first
19foundation · easy
Which of the following is an example of a Privacy-Enhancing Technology?
Select an answer first
20foundation · easy
What is the primary purpose of differential privacy in AI models?
Select an answer first

Finished these 5 questions?

Review the revealed explanations, or continue through the curriculum.

Next 5 questions →Back to CDPSE

Free Basic Practice is a study aid with revealable answers — not a scored exam. Examers.io is independent and not affiliated with or endorsed by ISACA. “CDPSE” is a trademark of its owner, used for identification only.