The correct answer is Variable group.

Variable group provides a centralized library for storing configuration values and secrets so multiple build jobs and release pipelines can reuse the same settings. You can link Variable group to several pipelines and update values in one place so all linked jobs pick up the changes automatically.

Variable group supports secret variables and can be referenced from YAML or classic pipelines which makes it appropriate for cross pipeline configuration management and consistent releases.

Service connection represents credentials and endpoints for external services and is used to authorize pipelines to access resources such as cloud subscriptions or artifact feeds. It does not serve as a general store for arbitrary build configuration values.

Task group bundles multiple pipeline tasks into a reusable unit to simplify pipeline definitions. It helps reuse steps but it is not designed to hold centralized configuration values for multiple builds or releases.

Full AWS Practitioner Certification Question