Sort transparency artefacts by who holds the pen and most of this domain sorts itself. A worked scenario runs the same three artefacts past an auditor; this card stays on the AWS-authored one. Alongside Amazon SageMaker Model Cards, the guide counts open source models, data and licensing among the tools for identifying transparent and explainable models, and those three need no document written at all. Explainability sits next door and is a different job: telling a person why a decision came out the way it did needs the reasoning behind one output, where a card describes the service in general.
Flash card
AWS AI Service Card: the transparency document AWS publishes for one of its own managed AI services or models.
- AWS writes it and AWS publishes it. You cannot author or edit one. The cards are public pages on the AWS documentation site, readable without an account, so an auditor or a regulator reads the same text you do.
- Each card covers one AWS AI service or model, or one capability inside a service, rather than a whole product family. Amazon Rekognition face matching has its own card; so does Amazon Textract AnalyzeID. One service can carry several cards, because the fairness and accuracy story differs by capability.
- The contents follow a set shape: an overview, the intended use cases and the limits AWS states, a design section on fairness, explainability, robustness, privacy and governance, then deployment best practices. Fairness carries measured numbers, such as match rates across demographic groups on named evaluation datasets.
- A service card documents behaviour, it does not change it. Reading one does not filter a response, gate a deployment or block a request; Amazon Bedrock Guardrails and IAM policies do that. Cite the card as evidence in a risk assessment or a procurement review.
- The counterpart for a model your own team builds is Amazon SageMaker Model Cards, which you author and version yourself: intended use, risk rating, training and evaluation details. An AI Service Card is AWS documenting AWS. A SageMaker Model Card is you documenting your model. AWS Artifact is neither, holding AWS’s downloadable SOC and ISO compliance reports.
Pick it when
Pick it when this scenario turns on what AWS has published about a managed service you call: what it is meant for, how its accuracy varies across groups, or what AWS recommends when deploying it. A document you wrote yourself proves nothing about a service somebody else runs.
It's the wrong answer when
Wrong for your own trained or fine-tuned model, where Amazon SageMaker Model Cards fit; for a compliance certification such as SOC 2 or ISO 27001, where AWS Artifact does; for measured scores over your prompts and data, where Amazon Bedrock evaluations do; and for a control acting at request time, where Guardrails or an IAM policy does.