Amazon Echo to Ditch Local Voice Processing for Cloud-Based Voice ID: Privacy Implications and User Choices

Amazon Discontinues Echo's Local Voice Processing in Favor of Voice ID

Amazon has announced that it will disable local voice request processing on its Echo smart speakers starting March 28. This change will require all voice commands to be processed through the cloud, effectively ending a key privacy feature. The move is part of Amazon's introduction of Alexa+ and Voice ID, which will enable Echo devices to recognize different users, allowing them to access personal information such as calendars, music libraries, and reminders.

While devices like Apple's Siri have been capable of recognizing multiple users since 2019, Siri required server connections for voice processing until iOS 15, which introduced on-device speech handling for tasks like setting alarms and controlling music. For current queries and system updates, internet access is still necessary across all smart speaker brands.

Amazon's decision to move away from local processing also affects the 'Don't save recordings' feature. Users will need to disable this feature by March 28 to ensure Voice ID works properly; otherwise, their Echo devices may become non-functional. The company has a history of privacy concerns, including employees listening to audio samples for training purposes and paying fines for Ring camera violations.

Amazon states that it will delete recordings after cloud processing, but users must disable the 'Don't save recordings' control on their local devices. This shift is seen as a step towards potential future subscriptions, as the Alexa service has been unprofitable for Amazon.