Samsung has placed a strong emphasis on user privacy with its recent software updates, offering dedicated settings intended to restrict artificial intelligence processing entirely to the device. However, new findings indicate that these safeguards may not cover every native feature as thoroughly as users might expect.
According to a report from Android Police, enabling Samsung's AI privacy toggle failed to stop Bixby from sending voice input to external servers. While the setting is designed to keep AI data processing strictly on-device, the voice assistant still routed audio data off the phone to handle user requests.
This behavior highlights the ongoing complexities of managing privacy controls in the era of smartphone AI. While broad toggles provide a layer of local processing assurance, specific apps and built-in assistants may continue operating under their own rules, leaving sensitive voice interactions dependent on the cloud.
For a complete breakdown of the testing and technical details, check the original coverage on Android Police.
