Data in their dual role as the representatives of the real world and as fuel for AI applications constitute the bridge between the two worlds of artificial intelligence and human intelligence. Data in medicine and healthcare are characterized by many peculiarities, such as limited access due to privacy and security demands, scarcity in case of rare diseases, but also by the lack of interoperability. The FAIR principles and standardization effort, such as the OMOP Common Data Model, offer solutions to manage data so that they can be used for AI developments. As data quality is key, health data governance policies have to be put into place. The future of robust and reliable AI in healthcare is coupled with dataware, the availability of very large health data spaces.