advertisement
A private conversation between a couple in the US was recorded by Alexa and then sent to a random person in their contact list.
Amazon on Thursday described an "unlikely... string of events" that made Alexa send an audio recording of the family to one of their contacts randomly. The episode underscored how Alexa can misinterpret conversation as a wake-up call and command.
Amazon’s defence of this incident is rather perplexing. The Jeff Bezos-owned giant simply highlighted the process in which Alexa functions across devices.
“Alexa, which comes with Echo speakers and other gadgets, starts recording after it hears its name or another "wake word" selected by users. This means that an utterance quite like Alexa, even from a TV commercial, can activate a device. That's what happened in the incident,” Amazon said.
Amazon has repeatedly claimed that its voice assistant service, Alexa whose data is recorded on the cloud, is safe and doesn’t violate privacy issues.
Incidents like this clearly highlight the implications of owning such devices, especially the ones like Google Home that doesn’t even have a microphone that can be stopped from hearing conversations.
University researchers from Berkeley and Georgetown found in a 2016 paper that sounds which are unintelligible to humans can set off voice assistants in general, raised concerns of exploitation by attackers.
At that time, Amazon did not comment on the matter, but it previously told The New York Times that it has taken steps to keep its devices secure.
Amazon and Google (biggest hoarders of user data) top the smart speaker market with majority of their devices selling in countries like the US and parts of Europe. Their next-big target is India, where data privacy regulations are yet to be enforced.
(At The Quint, we question everything. Play an active role in shaping our journalism by becoming a member today.)