Databricks Generative AI Engineer Associate Question 245
Select 3You are tasked with implementing governance for a Databricks workspace used to train generative AI models. Which of the following practices will help ensure secure model development and compliance with organizational policies?
- A
Enable Unity Catalog to manage data access and lineage tracking.
- B
Grant all users unrestricted access to the workspace to foster collaboration.
- C
Use MLflow to track model versions and associated metadata.
- D
Ensure that sensitive training data is encrypted both in transit and at rest.
- E
Disable audit logs to improve system performance during training.
Show answer and explanation
Correct answers: A, C, D
Explanation
Governance in generative AI development involves ensuring data security, tracking model versions, and implementing access controls. Unity Catalog, MLflow, and data encryption are critical tools for maintaining compliance and security. Conversely, unrestricted access and disabling audit logs violate governance best practices and expose the system to risks.
- A. Correct.
Unity Catalog provides centralized data access control and lineage tracking, ensuring proper governance over data used in AI model training.
- B. Incorrect.
Granting unrestricted access contradicts governance principles as it increases the risk of unauthorized access and data breaches.
- C. Correct.
MLflow enables model tracking and version control, which are crucial for governance and auditability in generative AI workflows.
- D. Correct.
Encrypting sensitive training data ensures compliance with data protection regulations and secures the data against unauthorized access.
- E. Incorrect.
Disabling audit logs undermines governance efforts, as audit logs are essential for monitoring, compliance, and tracing activities in the workspace.