Guidelines for Responsible AI

Official references for this objective

• Amazon Web Services — Considerations for addressing the core dimensions of responsible AI for Amazon Bedrock applications | Artificial Intelligence

  For toxicity, you can use either RealToxicityPrompts or BOLD datasets, or both.

• Amazon Web Services — Responsible AI Lens - AWS Well-Architected Framework - Responsible AI Lens

  Do not use this guidance as a compliance or assurance checklist.

• Amazon Web Services — Responsible AI – Building AI Responsibly – AWS

  safety protections that block up to 88% of harmful content and deliver auditable, mathematically verifiable explanations for validation decisions with 99% accuracy

• Amazon Web Services — Announcing the AWS Well-Architected Responsible AI Lens  | Artificial Intelligence

  The narrower the scope of the use case, the simpler the time you will have identifying, mitigating, and testing risks that the AI use case

• Amazon Web Services — Build responsible AI applications with Amazon Bedrock Guardrails | Artificial Intelligence

  You should still perform your own independent assessment and take measures to ensure that you comply with your own specific quality control practices and standards

• Amazon Web Services — Generative AI Data Governance – Amazon Bedrock Guardrails – AWS

  Automated Reasoning checks in Amazon Bedrock Guardrails is the first and only generative AI safeguard to use formal logic to help prevent factual errors from

• Amazon Web Services — Governance by design: The essential guide for successful AI scaling | Artificial Intelligence

  Using Amazon Bedrock, the Innovation Center conducted an AI-powered evaluation. This established transparent, data-driven risk management

• Amazon Web Services — Introducing AWS AI Service Cards: A new resource to enhance transparency and advance responsible AI | Artificial Intelligence

  we recommend an iterative approach where customers periodically test and evaluate their applications for accuracy or potential bias

• Amazon Web Services — Model Explainability - Amazon SageMaker AI

  Transparency about how ML models arrive at their predictions is also critical to consumers and regulators.

• Amazon Web Services — Pre-training Bias Metrics - Amazon SageMaker AI

  All of the pretraining metrics are model-agnostic because they do not depend on model outputs and so are valid for any model.