Google Professional Cloud Security Engineer Question 358
Select 3Google Cloud PlatformYou are deploying a machine learning model on Google Cloud AI Platform and need to ensure the workload is secure. Which of the following actions should you take to protect your AI workload from unauthorized access and data breaches?
- A
Enable IAM roles and permissions specific to AI Platform users and services.
- B
Store sensitive training datasets in a public Cloud Storage bucket for easy access.
- C
Use VPC Service Controls to define security perimeters around AI Platform services.
- D
Ensure training data is encrypted both at rest and in transit.
- E
Allow unrestricted access to the AI Platform endpoint for ease of integration.
Show answer and explanation
Correct answers: A, C, D
Explanation
Securing AI workloads on Google Cloud requires implementing best practices such as configuring IAM roles, encrypting sensitive data, and using network-level security controls like VPC Service Controls. Avoiding practices like storing data in public buckets or exposing endpoints without restrictions is critical to maintain the integrity and confidentiality of the workload.
- A. Correct.
Assigning specific IAM roles and permissions ensures that only authorized users and services can access and manage the AI Platform resources, reducing the risk of unauthorized access.
- B. Incorrect.
Storing sensitive training data in a public Cloud Storage bucket is a significant security risk, as it exposes the data to unauthorized access.
- C. Correct.
Using VPC Service Controls helps to create security perimeters around the AI Platform services, preventing unauthorized access from outside the defined boundaries.
- D. Correct.
Encrypting training data both at rest and in transit provides an additional layer of security, ensuring that the data cannot be easily intercepted or accessed by malicious actors.
- E. Incorrect.
Allowing unrestricted access to the AI Platform endpoint exposes your workload to potential unauthorized access and malicious activity, violating security best practices.