The correct choice is Amazon SageMaker Clarify. It is purpose built to detect and quantify bias in datasets and in model predictions and it generates explainability reports that teams can review before promoting a model to production.

Amazon SageMaker Clarify can compute fairness metrics on training data and on predicted outcomes and it can produce feature attributions to help explain which inputs drive decisions. It integrates with SageMaker training and pipeline workflows so bias and explainability checks can be automated as part of model development and pre deployment validation.

Amazon Comprehend is focused on natural language processing tasks such as entity recognition and sentiment analysis and it does not provide model fairness or bias assessment tools.

AWS Glue DataBrew helps with visual data preparation and cleaning and it lacks built in capabilities to compute model or dataset bias reports.

Amazon SageMaker Model Monitor is intended to watch deployed models for data drift and quality issues in production and it is not designed for the pre deployment bias evaluation that Clarify performs.

Full AWS Practitioner Certification Question