ISACA • CDPSE
Validates the technical skills and knowledge to assess, build and implement comprehensive data privacy measures across privacy governance, risk management, data lifecycle, and privacy engineering.
Questions
749
Duration
210 minutes
Passing Score
450/800
Difficulty
ProfessionalLast Updated
Jan 2026
The Certified Data Privacy Solutions Engineer (CDPSE) is a globally recognized, experience-based technical certification awarded by ISACA that validates the skills required to assess, build, and implement comprehensive data privacy measures. Unlike policy-focused privacy credentials, CDPSE is specifically designed for technology professionals who translate privacy requirements into working technical solutions — implementing privacy by design across systems, networks, and applications. The certification covers four core domains: Privacy Governance, Privacy Risk Management and Compliance, Data Life Cycle Management, and Privacy Engineering, with particular emphasis on technical implementation areas such as encryption, anonymization, identity and access management, and privacy-enhancing technologies (PETs).
First introduced by ISACA, the CDPSE has grown to more than 16,000 credential holders worldwide and was updated with a revised Body of Knowledge taking effect in April 2025, reflecting evolving regulations such as GDPR and CCPA, emerging AI/ML privacy challenges, and modern infrastructure requirements. The certification demonstrates that holders can not only understand privacy frameworks but engineer privacy controls into real-world technology platforms and data pipelines.
CDPSE is intended for mid-to-senior level technology professionals who are actively involved in building and implementing privacy solutions rather than defining policy. Relevant job roles include Privacy Engineers, Data Protection Engineers, Security Architects, Cloud Engineers, DevOps professionals with privacy responsibilities, IT Risk Managers, and Compliance Technologists. Professionals working in environments subject to GDPR, CCPA, HIPAA, or other data protection regulations will find particular value in this credential.
Candidates are expected to have a minimum of three years of cumulative work experience performing CDPSE job practice tasks within the ten-year period preceding their application. The exam itself is open to anyone, including those who have not yet met the experience threshold, but full certification requires verified professional experience submitted through an ISACA account within five years of passing the exam.
There are no formal educational prerequisites to sit for the CDPSE exam. However, ISACA recommends that candidates have at least three years of hands-on experience in roles involving privacy technology implementation, data governance, risk management, or security engineering. This experience must be directly tied to the four CDPSE job practice domains and verifiable by a supervisor or manager.
A solid foundational understanding of networking, cloud infrastructure, application development, and information security is strongly recommended before attempting the exam. Familiarity with major privacy regulations (GDPR, CCPA), Privacy Impact Assessments (PIAs), data classification methodologies, encryption standards, and identity and access management concepts will be essential. Professionals who already hold ISACA certifications such as CISA or CISM, or industry credentials such as CISSP or CIPP, will find significant content overlap and may require less preparation time.
The CDPSE exam consists of 120 multiple-choice questions, each with a single best answer, to be completed within 210 minutes (3.5 hours). Questions are scenario-based and assess applied knowledge rather than rote memorization, requiring candidates to evaluate real-world privacy engineering situations. The exam is scored on a scale of 200 to 800, with a passing score of 450. ISACA uses scaled scoring to account for variation in difficulty across exam versions.
The exam is delivered as a computer-based test and is available at authorized PSI testing centers worldwide or via remote proctoring, allowing candidates to test from their own location. Registration is open on a continuous basis, and testing appointments can be scheduled as early as 48 hours after fee payment. The exam is available in English, Chinese Simplified, Spanish, and German. Candidates who do not pass may retake the exam up to four times within a rolling 12-month period, with each attempt requiring full payment of the exam fee ($575 for ISACA members, $760 for non-members).
CDPSE-certified professionals are positioned at the intersection of two high-demand fields — cybersecurity and data privacy — making them highly sought after as organizations scale their compliance programs to meet GDPR, CCPA, and other global regulations. Common roles for credential holders include Privacy Engineer, Data Protection Officer (technical track), Security Architect, Cloud Privacy Specialist, and IT Risk Analyst with privacy focus. ISACA data indicates that the average annual salary for CDPSE holders in the United States exceeds $150,000, ranking it among the top-paid certifications in information security. More than half of credential holders report applying CDPSE skills daily, and 42% report measurable productivity gains attributable to the certification.
Compared to policy-oriented privacy credentials such as the IAPP's CIPP or CIPM, CDPSE occupies a distinct technical niche, making it the preferred credential for engineers and architects rather than privacy counsel or compliance officers. For professionals who already hold CISA, CISM, or CISSP, CDPSE adds a specialized privacy engineering layer that complements broader security governance credentials. With more than 16,000 holders globally and growing regulatory pressure across industries including healthcare, finance, and technology, demand for CDPSE-qualified professionals continues to increase.
5 sample questions with correct answers and explanations. Start a practice session to test yourself across all 749 questions.
1. A financial services consortium needs to detect money laundering patterns across three competing banks without revealing individual transaction details to other participants. The solution must provide privacy for each bank's sensitive data while enabling collaborative fraud detection. The security architecture must protect against malicious participants who might deviate from protocols. Which secure computation approach should the privacy engineer implement? (Select one!)
Explanation
SPDZ (pronounced 'Speedz') protocol provides practical secure multi-party computation with malicious security, meaning it protects against participants who arbitrarily deviate from the protocol - essential when competing banks collaborate. SPDZ uses additive secret sharing to split sensitive data so no single party learns others' inputs, and provides cryptographic guarantees even if participants actively attempt to cheat or extract information. The malicious security model is critical in financial contexts where competitive interests might motivate protocol violations. Yao's Garbled Circuits only provide semi-honest security, assuming participants follow the protocol honestly, which is insufficient for competing banks. Homomorphic encryption with separate keys prevents collaborative computation entirely since operations require consistent encryption schemes. Federated learning with local DP is designed for machine learning model training, not the exact pattern matching required for fraud detection across transaction databases.
2. A privacy team implements data subject rights fulfillment for Right to Erasure requests under GDPR Article 17. The organization's data architecture includes production databases with nightly backups retained for 90 days, a data warehouse with 7-year retention, and machine learning models trained on historical customer data. Technical challenges prevent immediate erasure from all systems. Which two approaches align with GDPR requirements for handling these technical limitations? (Select two!)
Multiple correct answersExplanation
Implementing logical deletion flags in production systems while allowing backups to expire naturally is acceptable under GDPR, as backups are not considered active processing and erasure from backups is not required if they will be deleted within their normal retention schedule. Maintaining documentation of technical limitations is required to demonstrate compliance efforts and explain why immediate erasure from certain systems (like ML models) is not technically feasible. Retraining all ML models immediately is often technically infeasible and disproportionate; model unlearning for large language models and complex models remains an unsolved technical challenge. Refusing erasure requests until backups expire violates Article 17 obligations to act without undue delay. Ignoring backups and analytics systems entirely without documentation or justification fails to demonstrate good-faith compliance efforts.
3. A California-based advertising technology company implements Global Privacy Control support for its real-time bidding platform. The engineering team detects the Sec-GPC HTTP header with value 1 and the JavaScript property navigator.globalPrivacyControl returns true. Under California law effective in 2025, what is the required business response when GPC signals are detected? (Select one!)
Explanation
California law requires frictionless opt-out under CCPA/CPRA, meaning when GPC signals are detected, businesses must immediately stop selling and sharing personal information for that user without confirmation dialogs, delays, or verification steps. The GPC signal itself constitutes a valid consumer request that must be honored in real-time. Displaying confirmation dialogs violates the frictionless requirement. While regulations effective January 2026 require confirmation that opt-out requests have been processed, the opt-out must take immediate effect upon detection. The 45-day response period applies to other consumer rights requests like access and deletion, not opt-out signals which must be processed immediately.
4. A privacy engineer implements differential privacy for a medical research database publishing aggregate statistics. The epsilon parameter is set to 0.8 for simple statistical queries. After running five independent queries using the Laplace mechanism, what is the total privacy budget consumed under the composition theorem? (Select one!)
Explanation
Under the composition theorem for differential privacy, sequential application of multiple mechanisms accumulates privacy loss additively. If each of five queries consumes epsilon of 0.8, the total privacy budget consumed is 5 times 0.8 equals 4.0 epsilon. This represents linear composition under basic differential privacy guarantees. Independence of queries does not prevent privacy budget accumulation. Averaging epsilon values is incorrect. The square root formula applies to advanced composition theorem with approximate differential privacy parameters, not basic epsilon composition.
5. A Data Protection Officer is determining whether a Data Protection Impact Assessment is mandatory for a new employee wellness program. The program will process health data and fitness tracker information for 8,000 employees, use automated algorithms to provide personalized health recommendations, and monitor employee activity patterns to suggest interventions. However, participation is voluntary and does not produce legal effects on employees. Which GDPR Article 35 criteria make a DPIA mandatory for this processing? (Select one!)
Explanation
Large-scale processing of special category health data is the mandatory DPIA trigger because the program processes health data and fitness information (Article 9 special categories) for 8,000 employees, which exceeds large-scale thresholds. According to Estonia DPA guidance, 5,000 individuals is the threshold for sensitive data processing, and this program significantly exceeds that with 8,000 employees. Systematic and extensive automated evaluation would require legal or similarly significant effects on individuals, but the scenario explicitly states participation is voluntary and does not produce legal effects, so this criterion is not met. Systematic monitoring of publicly accessible areas does not apply because employee wellness programs occur in private workplace contexts, not publicly accessible areas. Processing of criminal conviction data is not relevant because the scenario involves health and fitness data, not criminal records.
One-time access to this exam