Earlier this week, Amazon’s smart personal digital assistant Alexa created some buzz after news came out that she had secretly snooped on a Portland couple, whose private conversation she sent to the man’s colleague sitting all the way in Seattle. Hours after the report first came out, Amazon spokesperson confirmed the incident and gave a bizarre and rather turbid explanation of the entire episode.
“Echo woke up due to a word in background conversation sounding like ‘Alexa’. Then, the subsequent conversation was heard as a ‘send message’ request. At which point, Alexa said out loud ‘To whom?’ At which point, the background conversation was interpreted as a name in the customer’s contact list. Alexa then asked out loud, ‘[contact name], right?’ Alexa then interpreted background conversation as ‘right’,” the Amazon spokesperson said in the statement.
The interesting thing to note here is that entire explanation is based on a lot of assumptions. Assumptions about why Alexa acted the way she did. And it if you carefully notice, with this incident, Amazon is trying to draw a line in the dark. Clearly, either the company has no idea of what is going on or that statement is just meant to pacify the affected couple and prevent public outrage.
The question here, however, is not limited to just Amazon Echo and Alexa. In a way it is about all the internet connected devices and the information that these devices sent to various websites and companies, but of particular note are the smart speakers that are supposed to be on your bedside table and are supposed to be the kind of gadgets with which you talk. These are personal in a way even our phones are not. And yet everything smart speakers do involve them listening to your voice and sending your words to the powerful computers inside data centres run by Silicon Valley companies. The question here is: By bringing a smart speaker like the Amazon Echo or the Google Home, are you bringing a virtual – and benevolent – spy inside your bedroom?
Almost eight months back in October 2017, Google confirmed that a tiny glitch in its Google Home Mini smart speaker enabled the device to secretly record the users’ conversations without giving them the slightest hint of the recording. While the company did fix the bug with a software update, but that doesn’t entirely rule out the threat that Google Home Mini and other smart speakers like it pose to privacy.
So should you worried that your smart speaker is constantly snooping on you? The answer is probably no. But it’s also not quite that simple. The key lies in understanding the way these gadgets work.
Google Home and other smart speakers are listening to all your conversations all the time. They are actively looking for “hotwords” or “wake words” like “OK Google”, “Hey Google”, “Hey Alexa”. In theory, these wake words act as triggers for the device. These smart speakers record tiny snippets of your conversation and send it to the cloud to analyse and deliver specific response to the user. The recorded clippings are deleted from the cloud when the hotword isn’t detected.
Google Home and other smart speakers are listening to all your conversations all the time
Amazon Echo devices work in a similar way, except they use a technique called wake word-spotting to detect the wake word and the Echo device streams the audio, including a fraction of a second of audio before the wake word, to the cloud only when it detects the wake word.
This means that Google does not necessarily need your permission for collecting data about you as by using its services you are already giving your consent to its terms and conditions. This also indicates two things– while all the user data might be safe in the company’s vaults, it is not negating its presence and it doesn’t guarantee that if hackers haven’t cracked into Google’s servers yet, they won’t be able to do that in the future. To be fair, such a guarantee no company can give.
As far as Amazon Echo devices are concerned, Amazon on its web page for Alexa and Alexa Devices states, “Alexa uses your voice recordings and other information, including from third-party services, to answer your questions, fulfill your requests, and improve your experience and our services.” All your conversations with Alexa are being recorded and while there is a way of deleting all the recordings, the meaning gets lost in the translation.
Amazon states, “Deleting voice recordings may degrade your Alexa experience. If you delete voice recordings, we will also remove Home Screen Cards in the Alexa App related to those voice recordings. However, removing a Home Screen Card from your Alexa App does not delete voice recordings, adding, “Note that deleting your voice recordings does not remove your Alexa messages.” To put it simply, when it comes to Alexa, things happen at Amazon’s behest.
Data privacy is a serious matter and while these tech giants might be taking every precaution in the book to keep our personal information under layers of encryption, the problem is classic: Future changes. The data that is collected right now for some purposes can be used for something else in future. What if the company policies change in future? What if government regulatory environment change in future? What if there is a security breach and hackers download all the saved conversations and then use them to target people?
All your conversations with Alexa are being recorded and while there is a way of deleting all the recordings, the meaning gets lost in the translation
Apart from the risk of our private information getting leaked online, there is always a chance that internal bugs and glitches could pave way for our private conversations to land in someone else’s mailbox. And then there is always snooping on part of the governments. It is possible that some day governments may decide that the data smart speakers have collected is something it needs to access. Companies like Google and Amazon can fight governments but they can’t win. If they have data and any government wants it, they will have to give it this way or that way.
NOT ALL DOOM AND GLOOM
So what’s the way out for users? One suggestion that we have is this: Use them less and use them with caution. Google and Amazon too offer some way out. Amazon has a mute button which physically disables the internal circuitry, prohibiting the sensors from listening and watching anything. In theory, this means that once users press the mute button, the Echo device should be deaf and dumb. Google Home devices work in a similar way. Their microphone can also be muted although Google hasn’t specified if its device too physically disable the microphone circuitry when mute button is pressed.
But caution is equally important. While these companies may be giving us options to mute these devices, don’t take anything at face value. We all know how many times Facebook has apologised for making same privacy mistakes again and again. Google recently came under scanner in Australia after software giant Oracle, which by the way is fighting a nasty court battle with Google, in its report to Australian watchdog, Australian Competition and Consumer Commission (ACCC) pointed out that Google was collecting data from Android users even when the location services were turned off and there was no SIM card in the device.
The important thing to remember here is that these gadgets are as powerful as you let them be. The more services you access using these smart speakers, the more rights they have to snoop in your life. While using them sparingly might not necessarily guarantee you all the privacy that you need, but that sure restricts their access to a certain extent-which by all means is better than letting them run all over your personal lives.