Edge computing
Processing data where it originates — instead of sending it to a data centre.
Edge computing means moving computation to the edge of the network: onto the device where the data arises, rather than a server somewhere. The opposite model is the cloud, where everything is uploaded first and processed afterwards.
The advantages are practical: no waiting for uploads, no dependence on the connection, no running costs per processed image.
The most important point, though, is legal. What never leaves the device needs no data-processing agreement, no review of server locations and no debate about retention periods with a service provider.