Smart speakers such as Amazon’s ‘Echo’ or Google’s ‘Home’ ranges have grown in popularity massively recently, especially over the Christmas period. Featuring smart assistants such as Amazon’s ‘Alexa’, the speakers can answer your questions, call and message contacts, keep track of your schedule and reminders and play audio media on demand, along with a multitude of other capabilities.
These smart speaker devices are activated with a spoken ‘wake’ word or phrase. When activated, the speaker begins recording to the user’s question or instruction then the data is sent over the Internet to the speaker’s respective manufacturer’s online systems (Google, Amazon, etc.) where the data is analysed. The processed response is then sent back to the device where the request is followed. This process only takes a few seconds from start to finish and can be extremely convenient since no button pressing or mobile devices are required in the process. The speaker can be placed in a high traffic area of the home (such as the kitchen) and makes for an extremely convenient home assistant.
Despite the convenience, many consumers are worried that the speakers are listening in on conversations, even without being activated by the wake word. Many are also worried about the increasing number of household items that require Internet connections. The speakers could potentially be used without consent to collect huge amounts of information on an individual to build a digital profile, which would then be sold as marketing data to advertisers.
Users do not want the speakers to be eavesdropping on private conversations without consent. Speaker manufacturers Google and Amazon have reported their devices do not transmit voice data without the wake word being said. The Google Home Mini’s primary button that activates listening had a particular problem that resulted in the device recording constantly, without the user pressing it. Google solved the issue by disabling the button to assure the public that their smart devices were not constantly listening.
Recently, concerns were raised that a smart assistant’s data profile building technologies could be unintentionally gathering data from children. Sorin Toma, Principal Adviser on Cyber Security at the University of NSW and Managing Director of Xpotentia, a cybersecurity consulting firm, said “The first thing that parents need to understand when using the Alexa unit, is that whenever these devices are on, they are listening. When you consider that each unit has one account and that critically, the devices do not differentiate between adults and children, there is real cause for concern”.
Parents should be conscious of what features are activated while the device is in the presence of children. Features such as voice-activated purchases should be disabled, and parental controls enabled. Amazon’s Alexa recently added a plethora of parental features and controls to their line-up of Echo devices
It has not yet been proven that smart speaker technologies is being actively used to build data profiles on consumers without consent, though it should be considered that when the wake word is said, that the device is recording, processing and storing data on everything that is said.
Consumers should be conscious of smart technology and ensure that any sensitive information is not vocally presented after the wake word activates the device. Otherwise, consumers must rely on the word given by speaker manufacturers that their privacy is assured.