Edward Snowden previously warned about Google’s messaging app, Allo. The recent report suggests that Snowden’s claims were, in fact, correct, and Google’s new messaging application is not handy when it comes to day to day use.
Google’s new virtual assistant appears to have a bug that provided unrelated personal information to the user’s friends. According to Independent, company’s new app, Allo, is capable of sharing the user’s past Google searches with their contacts. According to the publication, the recent discovery was made by Re/code’s Tess Townsend, who apparently was using Google’s new messaging application to chat with a friend.
“In the middle of our conversation, my friend directed Assistant to identify itself,” wrote Ms. Townsend. “Instead of offering a name or a pithy retort, it responded with a link from Harry Potter fan website Pottermore.”
Ms. Townsend further stated that neither she nor her friend had mentioned Harry Potter in their conversation before Google’s new messaging application brought it up.
“But the response was not merely a non-sequitur,” continues Ms. Townsend. “It was a result related to previous searches my friend said he had done a few days earlier.”
Like any other messaging application, even Google’s Allo is supposed to request permission from a user before sharing their personal information on social media platforms or among their friends, but this privacy feature of Google does not always work. Things became awkward for Ms. Townsend when she asked Allo about her occupation. Instead of showing the publicly listed address of her employer, Google’s virtual assistant shared a Google Maps image with the address of a co-working space she had used in the past.
“Google had the address on file because I had included it in my personal Google Maps settings,” wrote Ms. Townsend. “It did not ask my permission to share that.”
It was not long ago when Edward Snowden issued a warning not to use Google’s messaging app under any circumstances. The famous whistleblower warned everyone not to use Allo as it lets the company read everything that the user say.
— Edward Snowden (@Snowden) September 21, 2016
Google’s virtual assistant comes with a robot that watches everything people say, and then it stores it for later analysis. The stored data helps the application to respond faster in the future. But it also means that all the chats may stay on Google’s servers indefinitely, and they are able to read it as well. Google had initially indicated that the stored messages would stay on the servers temporarily, limiting the possibility of any data breach.
Google had revealed in the past that by storing the information, it makes Allo more useful, but many people have claimed that by collecting that information, Google was putting people at risk of being spied on by the government.
Google has issued a statement about the slip-up and stated that they had fixed the issue.
“We were notified about the Assistant in group chats not working as intended. We’ve fixed the issue and appreciate the report.”
Meanwhile, Amazon’s smart gadget is also under the scrutiny. Earlier this month, a clip surfaced online that showed an anonymous woman asking Amazon’s Alexa a series of questions starting with, “would you lie to me” and finishing it with, “Alexa, are you connected to the CIA?”
Amazon’s virtual assistant swiftly responded to the earlier questions, but it abruptly shut down after the woman asked about its connections with the U.S. government agency.
— Stephen Janiszak (@StephenJaniszak) March 12, 2017
Amazon’s spokesperson told DailyMail.com that when the woman in the video asked those questions to their virtual assistant, it encountered with a glitch, and the spokesperson stated that Alexa’s response to “Are you connected to the CIA?” is “No, I work for Amazon.”
As of now, both the major companies claim to have resolved the ongoing issues.
[Featured Image by Spencer Platt/Getty Images]