AWS Certified AI Practitioner

Explanation

Correct Answer: B - Use data-encryption, guardrails, and human review of AI-generated outputs

Why this is correct:

1. Data Encryption: Medical imaging data like MRI scans contain sensitive patient health information (PHI) that requires protection under regulations like HIPAA. Encryption ensures data security both at rest and in transit.

2. Guardrails: AWS provides various guardrails and safety mechanisms to ensure AI systems operate within defined boundaries, preventing harmful or biased outputs.

3. Human Review: For medical applications, human oversight is crucial to validate the quality and appropriateness of synthetic data before using it for training models that could impact patient care.

Why other options are incorrect:

• A) Disable logging to protect privacy: While privacy is important, disabling logging removes audit trails and accountability mechanisms that are essential for responsible AI governance and compliance.

• C) Fully automate model approval with no oversight: Complete automation without human oversight is risky, especially in healthcare where decisions can impact patient outcomes. Responsible AI requires human-in-the-loop validation.

• D) Publish all synthetic data publicly for open research: Medical data, even synthetic, may contain patterns that could potentially reveal sensitive information about the original training data or patient populations. Responsible AI practices require careful consideration of data sharing risks.

AWS Best Practices for Responsible AI in Healthcare:

• Implement data governance and access controls
• Use AWS services like Amazon SageMaker with built-in security features
• Apply AWS AI Service Cards for transparency
• Follow the AWS Well-Architected Framework for AI/ML workloads
• Comply with healthcare regulations (HIPAA, GDPR, etc.)