Skip to content
Mementropy

14 — Implement AI Governance, Compliance, and Transparency

14 — Implement AI Governance, Compliance, and Transparency

Section titled “14 — Implement AI Governance, Compliance, and Transparency”

Module Introduction

Section titled “Module Introduction”
  • Welcome
  • Security, Governance, and Compliance: Three interconnected Concepts

Understanding Compliance Frameworks For Gen Ai Applications

Section titled “Understanding Compliance Frameworks For Gen Ai Applications”
  • Introduction
  • Governance and Compliance Strategy
  • Establishing a Governance Framework
  • Understanding Compliance Considerations for Gen AI Applications
  • AWS compliance program

Amazon Sagemaker Model Cards For Compliance Documentation

Section titled “Amazon Sagemaker Model Cards For Compliance Documentation”
  • Introduction
  • Amazon SageMaker Model Cards
  • Implementing SageMaker Model Cards for compliance documentation
  • SageMaker Model Card JSON Schema for GenAI Compliance
  • Model Customization with Amazon SageMaker AI
  • Enhanced Model Customization with SageMaker AI

Aws Glue For Automated Data Lineage Tracking

Section titled “Aws Glue For Automated Data Lineage Tracking”
  • Introduction
  • How AWS Glue Catalog Works
  • Supporting compliance and governance workflow integration
  • Implementation process

Metadata Tagging For Source Attribution In Fm Generated Content

Section titled “Metadata Tagging For Source Attribution In Fm Generated Content”
  • Introduction
  • Purpose and benefits
  • Key components
  • Use cases
  • Enhanced Source Attribution with Amazon Bedrock Knowledge Bases

Aws Cloudtrail Audit Logging To Maintain Traceability

Section titled “Aws Cloudtrail Audit Logging To Maintain Traceability”
  • Introduction
  • Auditing generative AI workloads with AWS CloudTrail
  • How it helps
  • The process for auditing generative AI workloads with AWS CloudTrail
  • Enhanced Security Monitoring with AWS Security Hub

Data Governance Strategies

Section titled “Data Governance Strategies”
  • Introduction
  • Data governance strategies for AI and generative workloads
  • Additional Resources for Data Governance with Generative AI Applications

Approaches For Implementing Governance Strategies

Section titled “Approaches For Implementing Governance Strategies”
  • Introduction
  • Governance strategies
  • Governance for securing generative AI
  • Monitoring AI systems

Best Practices For Ai Compliance Framework Implementation

Section titled “Best Practices For Ai Compliance Framework Implementation”
  • Introduction
  • Best Practices

Reasoning Checks For User-facing Explanations

Section titled “Reasoning Checks For User-facing Explanations”
  • Introduction
  • Automated reasoning checks in Amazon Bedrock Guardrails
  • Key features of automated reasoning checks
  • Implementing automated reasoning checks
  • Evidence presentation for source attribution
  • Amazon Bedrock AgentCore

Fairness Evaluation

Section titled “Fairness Evaluation”
  • Introduction
  • Pre-defined fairness metrics in CloudWatch
  • Setting up pre-defined metrics in CloudWatch
  • Amazon Bedrock Prompt Management for systematic A/B testing
  • Implementing systematic A/B testing with Prompt Flows
  • LLM-as-a-judge for automated model evaluations
  • Setting up LLM-as-a-judge evaluation
  • Advanced MLOps on Amazon SageMaker AI

Policy-compliant Ai Systems

Section titled “Policy-compliant Ai Systems”
  • Introduction
  • Understanding Guardrail policies
  • Guardrail policy enhancements
  • Implementing Amazon Bedrock Guardrails
  • Using Bedrock Guardrails for Policy Enforcement, Amazon Kiro Powers, and Specialized Content for Agents

Module Summary

Section titled “Module Summary”
  • Recap and next steps
  • Resources