Also known as: local processing · client-side processing
On-device processing means performing computation locally on a user's own hardware rather than transmitting data to a remote server, so that the raw input never leaves the device.
The privacy benefit is genuine for the interception and server-retention threat models: data that is never transmitted cannot be captured in transit or retained indefinitely by a provider.
Two qualifications are commonly omitted from marketing. Data remaining on the device is still retained on the device, reachable through physical access or device-targeted legal process; and several features described as on-device fall back to cloud processing under conditions the user is not clearly told about.
Edge AI
Edge AI is the practice of running AI models directly on end-user devices such as phones and laptops rather than on remote servers, keeping input data local to the device.
Federated Learning
Federated learning is a technique for training a shared model across many devices without centralising the training data: each device computes updates locally, and only those model updates — not the underlying data — are sent to a server for aggregation.
Data Minimization
Data minimisation is the principle that an organisation should collect and retain only the personal data actually necessary for a stated purpose, and no more, for no longer than needed.
Metadata
Metadata is the information surrounding a communication rather than its content — who contacted whom, when, for how long, from where, and how frequently — which remains visible even when the content itself is encrypted.