Definition
The information that providers of high-risk AI systems must provide to deployers (Article 13). Includes intended purpose, known limitations, expected performance, instructions for human oversight, and maintenance requirements. Must be understandable for the deployer.