SY0-701 Question 220
Single answerGeneral data considerationsA healthcare company is preparing to move several datasets into a cloud analytics platform so data scientists can study patient trends. The security administrator must reduce legal and privacy risk while still allowing the analytics team to perform useful analysis. One dataset includes patient names, medical record numbers, diagnoses, treatment dates, ZIP codes, and ages. Which action is the BEST choice before sharing this dataset with the analytics team?
- A
Hash all patient names with MD5 and leave the rest of the dataset unchanged so analysts can still link records easily
- B
Tokenize or de-identify direct identifiers and apply data minimization so only the fields necessary for the analytics use case are shared
- C
Compress the dataset and store it in a separate cloud bucket with limited administrator access
- D
Convert the dataset to read-only format and apply a legal hold so records cannot be modified during analysis
Show answer and explanation
Correct answer: B
Explanation
The best answer is to de-identify or tokenize direct identifiers and apply data minimization before sharing the dataset. In Security+ general data considerations, organizations should classify data, limit collection and sharing to the minimum necessary, and protect sensitive information throughout its lifecycle. In a healthcare-related scenario, privacy risk is especially important because even if direct identifiers are removed, combinations of remaining fields can still create re-identification risk. Best practices commonly reflected in privacy frameworks and guidance, including NIST privacy engineering concepts and data minimization principles used across regulatory environments, emphasize reducing unnecessary exposure of sensitive data and using de-identification or pseudonymization techniques where appropriate. Access control, storage segregation, and preservation controls are useful supporting measures, but they do not replace minimizing and de-identifying the dataset before broader internal use.
- A. Incorrect.
This is incorrect. While hashing is sometimes used to obscure values, MD5 is deprecated for security-sensitive purposes due to collision weaknesses, and hashing only patient names does not adequately address the privacy risk of the remaining data. Medical record numbers and combinations of quasi-identifiers such as ZIP code, age, diagnosis, and treatment date could still enable re-identification. Someone might choose this option because it appears to pseudonymize the data, but it does not sufficiently support privacy-focused data handling.
- B. Correct.
This is correct. Tokenization or de-identification of direct identifiers reduces exposure of sensitive personal information, and data minimization ensures the organization shares only what is required for the stated business purpose. This aligns with common security and privacy best practices for handling sensitive data, especially regulated health-related information. If analysts do not need names or medical record numbers, those fields should not be provided. Reducing unnecessary data elements also lowers the impact of accidental disclosure.
- C. Incorrect.
This is incorrect. Compression and separate storage may help with organization or access control, but they do not address the core issue of limiting exposure of sensitive data before it is shared for analysis. Even with restricted access, the analytics team would still receive more identifying information than necessary. This option focuses on storage location rather than appropriate data handling and privacy protection.
- D. Incorrect.
This is incorrect. Read-only permissions and legal hold relate to integrity, preservation, and e-discovery requirements, not privacy reduction for analytics use. A legal hold is typically used when data must be preserved for litigation or investigation, not as a primary method for protecting personal information in shared datasets. Someone might pick this if they confuse retention controls with privacy controls.