Google is introducing some adjustments to its voice assistant’s responsiveness. A blog post outlined new tweaks to setting up Voice Match with Google Assistant, as well as fine-tuning how likely it is to be activated.
The changes to activation are the more striking development. This rollout will let users determine how sensitive Google Assistant is to audio that sounds like its wake phrase. Most of the large tech companies with voice efforts, including Google, faced some privacy outcry when whistleblowers revealed that third-party reviewers might hear personal information in recordings. Google introduced new policies for reviewing voice recordings as a result, but issues remained. Those scandals also hinted at the scope of how often voice assistants are erroneously activated, with the AI accidentally waking up and recording audio of private work calls, doctor’s appointments, and other intimate activities.
This new privacy control is a step toward addressing those concerns. Options to adjust Google Assistant’s sensitivity will roll out in the coming weeks. The settings can be changed at any time, and can be set for each device running the voice assistant to better reflect its surroundings and use cases. For now, the feature is only available in English, but will have more languages added.
The other Google Assistant update is about recognizing individual people with Voice Match. This teaches the AI to differentiate between human voices, which means it can deliver more personalized results to queries. Voice Match lets up to six people be linked to a single device.