With a view to ensuring a high level of protection of fundamental rights, health and safety when AI systems are employed, it is necessary for high-risk AI systems to draw on high quality data, to have in place documentation of the data and to include appropriate data governance and management practices.