PMI-CPMAI repeatedly highlights data privacy and regulatory compliance as core elements of responsible AI, particularly when operational data, trade secrets, or other sensitive information is involved. A key technique recommended in responsible data handling is data anonymization or de-identification, which reduces the risk of sensitive details being exposed while still allowing AI models to learn useful patterns.
From a governance and compliance standpoint, anonymization supports principles such as data minimization and privacy-by-design, both of which are prominent in modern regulatory regimes. Even when the data is not strictly "personal," sensitive operational data can present competitive, security, or safety risks if improperly exposed. Anonymization can involve removing or masking identifiers, aggregating data, and transforming features so that individual entities or critical operational specifics cannot be reverse-engineered, while preserving statistical utility for modeling.
Zero-trust architectures and encryption schemes (options A and D) are important security controls, but they focus primarily on controlling access and protecting data in transit or at rest, not on reducing identifiability of the data itself. Secure multiparty computation (option B) is specialized and often beyond what is pragmatically needed for typical operationalization scenarios. PMI-CPMAI's responsible AI practices emphasize anonymization as a direct and effective privacy technique. Therefore, applying data anonymization to the dataset (option C) is the most appropriate choice.

The best answer is A. Adopt a rule-based extraction, transformation, and loading (ETL) framework . In PMI-CPMAI, the Identify Data Needs domain includes overseeing data cleaning, preprocessing, transformation, and validation so that data is suitable for model development. PMI's official exam outline specifically calls out defining data requirements, coordinating data cleaning and normalization, verifying preprocessing results, and ensuring the prepared data meets the format and quality needed for the intended AI approach.
An ETL framework is the most direct fit because the scenario is about getting telemetry data into the right format and shape for model use. ETL handles extraction from source systems, transformation into a usable model-ready structure, and loading into the target environment in a controlled, repeatable way. By contrast, DDS is more about data exchange architecture, not primary preparation for modeling. ML algorithms are used to learn from prepared data, not to format it. Batch processing may improve throughput, but performance optimization does not solve the core requirement of structuring and transforming the data correctly. Under PMI-CPMAI logic, data preparation should be systematic, auditable, and aligned to the model's requirements, which makes ETL the strongest answer.

The best answer is C. Utilize a diverse set of test cases . PMI-CPMAI's model evaluation domain focuses on building comprehensive evaluation plans and formulating appropriate evaluation questions and criteria.
Validation is not treated as a one-time technical check, but as a structured process designed to test model behavior across a range of relevant conditions, edge cases, and data contexts. Using a diverse set of test cases is the best way to assess whether predictions are accurate, robust, and dependable enough for a public-sector policy setting.
Option A is useful for software quality but does not validate predictive performance. Option B is weaker because a single validation exercise can miss important failure modes, bias, or context-specific weaknesses.
Option D supports engineering discipline, but continuous integration testing focuses more on code and deployment workflow than on validating model prediction quality itself. PMI's CPMAI framework emphasizes comprehensive evaluation design, iteration, and addressing performance issues such as drift and changing conditions. That makes broad and varied test coverage the most PMI-aligned approach to model validation. In practical terms, diverse test cases provide stronger evidence that the model will generalize beyond a narrow sample and support trustworthy decision-making.

The best answer is D. Implement a robust data validation and correction process . In PMI-CPMAI, data understanding and data preparation require the team to evaluate training data requirements, validate data quality, perform data cleansing and enhancement, and make go/no-go decisions based on whether the data is fit for model development. When discrepancies are detected during collection, the first priority is to validate the data, identify the source of the inconsistency, and correct or isolate bad records before moving further into integration or modeling.
Option A may eventually be necessary, especially when combining sensor, log, and database sources, but harmonizing formats should not come before confirming whether the sensor data is accurate and reliable.
Option B is not a first-step governance response and does not directly address the quality issue. Option C could be appropriate only if the validation process shows that the sensors themselves are faulty; replacing hardware before confirming the root cause would be premature. PMI's methodology consistently stresses data quality validation and cleansing as foundational activities in AI projects. Since the scenario explicitly mentions discrepancies, the most appropriate first action is to validate and correct the data so later integration and model-building decisions are based on trustworthy inputs.

The best answer is A. Using generative AI (GenAI) to create additional relevant data . PMI's official CPMAI exam content outline specifically includes supervising data augmentation and synthetic data generation within the data-preparation responsibilities of an AI project professional. That makes this choice the clearest PMI-aligned answer when the goal is to increase data quantity in a controlled way because information is missing or insufficient. Generative AI can help create additional relevant synthetic examples that support model development, provided the team also validates quality, documents transformations, and manages bias carefully.
The other options do not directly address the stated objective. Responsible AI techniques are important for governance and ethics, but they do not themselves augment the data set. Rule-based filtering may clean or reduce data, not increase it. Sentiment analysis is a modeling technique for a particular kind of text problem and is unrelated to filling data shortages in general. PMI's broader trustworthy AI guidance also stresses that synthetic or augmented data must be handled responsibly so that teams do not introduce new distortions while trying to solve a quantity problem. That is why GenAI-based creation of additional relevant data is the strongest answer, as long as it is paired with validation and bias controls.