Using AEAD Functions for Sensitive Data Protection in BigQuery

PDE By islever Editorial Last updated

Short answer

D Use AEAD functions and delete keys when employees leave the organization to ensure sensitive data becomes unreadable.

Question

Your organization stores highly personal data in BigQuery and needs to comply with strict data privacy regulations. You need to ensure that sensitive data values are rendered unreadable whenever an employee leaves the organization. What should you do?
  1. Use column-level access controls with policy tags and revoke viewer permissions when employees leave the organization.
  2. Use dynamic data masking and revoke viewer permissions when employees leave the organization.
  3. Use customer-managed encryption keys (CMEK) and delete keys when employees leave the organization.
  4. Use AEAD functions and delete keys when employees leave the organization. Correct Answer

Key Takeaways

Tests the difference between access control (IAM/Policy Tags) and cryptographic data protection (AEAD), where key deletion provides permanent unreadability.

Learn why AEAD functions are the correct solution for ensuring sensitive data in BigQuery becomes unreadable when employees leave, compared to IAM controls or CMEK.

Common mistake: Choosing CMEK (Option C) because it involves key management, but failing to realize it encrypts entire datasets rather than allowing per-user granularity needed for individual departures.

Expert Analysis

The Requirement for Immediate Unreadability

The scenario demands that sensitive data become unreadable the moment an employee leaves. This is a critical distinction from standard access control mechanisms, which rely on permission revocation. While IAM changes are necessary, they do not inherently encrypt or obscure data already in transit or cached by the user's session. The solution must address the data itself.

Why AEAD is the Correct Mechanism

Option D leverages BigQuery’s native support for Authenticated Encryption with Associated Data (AEAD) SQL functions. These functions allow encryption and decryption of individual cell values using customer-managed keys. When an employee departs, deleting their specific cryptographic key renders the encrypted cells permanently unreadable to anyone lacking that key, including system administrators. This provides a technical guarantee of confidentiality that goes beyond logical access controls.

Limitations of Other Options

  • Option A (Policy Tags): Column-level access controls via policy tags restrict who can query sensitive columns based on IAM roles. However, if an employee retains access before leaving, or if data is exported, this does not make the data unreadable upon departure; it only prevents future queries by unauthorized users. It relies on persistent identity management rather than cryptographic destruction.
  • Option B (Dynamic Data Masking): Dynamic masking alters how data appears in query results based on user permissions. Like policy tags, it is a visibility layer, not a storage-layer encryption mechanism. Deleting viewer permissions stops access, but it does not encrypt the underlying stored values, nor does it provide the same level of cryptographic assurance as key deletion.
  • Option C (CMEK): Customer-Managed Encryption Keys protect data at rest at the table or dataset level. Deleting a CMEK would render all data in that scope unreadable, which is disproportionate and disruptive to other employees' work. It lacks the granular, per-user granularity required to handle individual departures without impacting organizational operations.

Conclusion

Only AEAD allows for fine-grained, per-cell encryption tied to specific user identities. Destroying the key associated with a departing employee ensures their sensitive data is cryptographically erased, satisfying the strict privacy regulation requirement.

Exam Strategy

When questions specify 'unreadable' or 'cryptographic protection' triggered by a user event, prioritize options involving key deletion or token invalidation over permission revocation. Always check if the encryption scope (table-level vs. cell-level) matches the granularity required by the scenario.

Frequently Asked Questions

Why not use CMEK for employee data privacy?

CMEK encrypts data at the table or dataset level. Deleting a CMEK affects all data under that key, disrupting other users. AEAD allows per-cell encryption tied to specific user keys.

Does dynamic data masking make data unreadable?

No. Dynamic masking hides data in query results based on permissions. The underlying data remains readable to authorized users and accessible if exported, unlike AEAD encryption.

More PDE FAQ →

Community Votes

D
100%

100% of anonymous learners picked answer D. Votes are pick records left by other test-takers — they are not the verified answer.

Discussion 3 contributions

Compiled from public community threads and reader submissions, lightly edited for clarity. Handles and vote tallies are shown as recorded, and comments reflect the views of individual contributors.

From public community threads 3

  1. DerfelCardarn 👍 3 Answered D
    C is too complex
  2. StefG_ITA 👍 1 Answered D
    AEAD (Authenticated Encryption with Associated Data): BigQuery supports AEAD SQL functions that allow you to encrypt/decrypt values ​​at the individual cell level. You can associate a key with an employee, and if they leave, you destroy their key. Granularity: Unlike other solutions, this one acts on the data itself, making it unreadable to anyone, regardless of their IAM permissions.
  3. AwesomeGCP 👍 1 Answered D
    The correct answer is D. Use AEAD functions and delete keys when employees leave the organization.

Reader comments & corrections

No reader corrections yet — spotted an error or have a note? Share it below.

Log in to comment, report an error, or add a note about this question.

← Back to PDE Study Guide