Apache Beam is an open-source programming model for batch and streaming data processing. Developers define a pipeline once, in Python, Java, Go, TypeScript, Scala, YAML or SQL, and run it on an execution engine called a runner, such as Apache Flink, Apache Spark or Google Cloud Dataflow. This keeps pipelines portable between engines and reduces dependence on one processing platform. Beam is software that you run yourself or through a managed runner, not a hosted service.
Beam is developed in the open by the Apache Software Foundation, a US non-profit (501(c)(3)) whose projects use the Apache License 2.0. Because no vendor operates a Beam service, no Beam provider holds your data: where it is processed depends on the runner and infrastructure you choose. A European organisation can run Beam on its own servers or with an EU provider. If you choose Google Cloud Dataflow as the runner, processing falls under that provider's terms and jurisdiction.