Access Amazon S3 data files directly using AWS Lake Formation permissions

This article announces Lake Formation's new capability to access S3 data files directly using Lake Formation permissions, eliminating separate S3 bucket policies.

• Data scientists can read/write S3 files using Spark APIs governed by Lake Formation permissions
• New GetTemporaryDataLocationCredentials() API vends temporary scoped credentials for S3 locations
• Unified permissions eliminate operational overhead and permission drift between systems
• Java plugin integrated into EMR 7.13+ automatically handles credential vending for S3 access
• Supports EMR Spark, SageMaker Unified Studio, and custom applications
• Single audit trail in CloudTrail for both table and file-level access
• Reduces complexity for operations teams managing separate permission policies

Lake Formation now provides unified governance for both SQL table access and direct S3 file access, simplifying data lake management while maintaining fine-grained access control and compliance auditing.