Google Home usually only responds to its active “wake word” – usually “Ok, Google,” “Hey, Google” or similar phrases. However, a supposed “accidental” change that came as part of a recent software update enabled the devices to keep listening without the need for the wake word.
The change was first noticed by a user on Reddit who spotted a notification on his phone from his Home alerting him to the fact that his smoke alarm was going off while he was cooking. As the smoke alarm in question was not a smart device, he realized that his Home device was listening to everyday sounds, and recognized the smoke alarm, despite him not using the wake word.
Other users also started reporting similar incidents. Some got alerts for the sound of glass breaking, popped bubble wrap, air compressor tanks and other noises that sounded like alarms.
Following these reports, a Google representative acknowledged to Protocol that the feature was “accidentally” enabled as part of a recent software update. The updates, the representative claimed, has since been rolled back.
Although the “accidentally” activated feature provides for greater security – as the case of the original Reddit poster shows, it could have detected an actual fire without the need for smart smoke detectors – it does so with the trade-off of less privacy.
Google isn't the only company whose products have this feature. In 2018, Amazon added a similar feature to its Echo speaker called Alexa Guard. In addition to listening for suspect noises, Alexa Guard also listens to see if people are whispering, which will then cause it to respond more quietly.
Unlike the "accidental" feature in Google Home, however, Alexa Guard is not on all the time. Rather, users must enable it via voice before leaving their houses.
While Google rolled the update back, the fact that it happened has troubling implications in light of its recent investments.
The company recently made a $450 million investment in home security provider ADT. With this, the company could possibly leverage the millions of smart speakers already in people's homes to become a home security superpower.
Once the deal closes, ADT's more than 20,000 installers will start selling Google-made security cameras, smart displays and other hardware, while ADT itself will integrate its own hardware more closely with Google's technology. (Related: Google rolling out “Orwellian nightmare” tech that uses spy cameras to watch you in your own home.)
“The goal is to give customers fewer false alarms, more ways to receive alarm events, and better detection of potential incidents inside and around the home,” wrote Rishi Chandra, Google Nest vice president and general manager, in a blog post.
Even before its investment in ADT, Google's plans for entering the home security market were plain to see. The company has been slowly adding home security features to its Nest ecosystem, culminating with the release of Nest Secure in November of 2017.
Google's actions when it comes to how these devices listen in on their users, however, seem to indicate that the latter may unknowingly be trading their privacy for security.
The recent incident with the accidental update isn't the first time that Google has enabled its devices to listen in on customers without them knowing about it. In 2019, reports came out that Google had installed hidden microphones in the Nest Secure alarm system without telling customers that they were there.
According to Google, the microphones were “never intended to be a secret,” but was only left off the product's box and web page because of an “error.” Users only found out that Google could have been potentially listening to them when the company announced that an update would allow the security system to use Google Assistant.
Learn more about how Big Tech may be spying on you at Surveillance.news.