Facebook is again stirring concerns about user privacy.
The social-media giant confirmed on Tuesday that it has been collecting audio from its users’ private conversations on the Messenger app and paying hundreds of outside contractors to transcribe them.
Facebook claimed to have only recently halted the practice. “Much like Apple and Google, we paused human review of audio more than a week ago,” a company representative said.
According to Facebook, the users affected were those who chose to have their audio transcribed into text within the app. The social network claimed they used human transcribers in order to review the quality of their in-app transcription AI.
One of the known firms performing the transcriptions is the Santa Monica, California-based TaskUs, which also helps screen political ads for Facebook.
TaskUs employees aren’t allowed to publicly disclose whom they work for and must instead use a code name to refer to clients.
TaskUs employees who were uncomfortable with the work revealed the nature of their transcription assignments — on the basis of anonymity for fear of losing their jobs. They reported hearing Facebook users’ personal conversations, which sometimes included vulgar language. However, they did not know why they were being asked to transcribe it.
TaskUs echoed Facebook’s claim that the transcription work has been suspended. “Facebook asked TaskUs to pause this work over a week ago, and it did,” the company said.
Amazon, Apple, and Google have also come under fire for submitting their users’ audio to human review, a practice that critics say constitutes a violation of privacy.
In April, it was reported that Amazon had thousands of workers around the world listening to user requests to its Alexa virtual assistant — with the ostensible aim of improving the software. Apple’s Siri and Google’s Google Assistant similarly used human review, though all three tech companies now say they no longer do so.
Facebook has not disclosed to users that their audio may be disclosed to third parties, though its terms do state that it can collect “content, communications and other information you provide” when users “message or communicate with others.”
Facebook’s terms also state that its “systems automatically process content and communications you and others provide to analyze context and what’s in them,” although no mention is made of other human beings screening Messenger content.
Facebook founder and CEO Mark Zuckerberg has denied that Facebook uses the contents of Messenger conversations in determining the ads people see in their feeds.
“You’re talking about this conspiracy theory that gets passed around that we listen to what’s going on on your microphone and use that for ads,” he told the Senate during a congressional testimony last year. “We don’t do that.”
But the sharing of Messenger conversations with third parties raises important privacy concerns, even if the content is not used for advertising purposes.
For example, will private conversations in Messenger eventually be used to determine whether a user is violating “hate speech” standards — and thus used to justify censoring or banning users even if their public content follows community standards?
And will the private comments of users be “leaked” by third-party reviewers with opposing political views in order to disgrace them and stoke intimidation?
In response to concerns about spying, Facebook claims its app only accesses users’ microphones if they are using a specific feature that requires audio, such as voice messaging.
Can users take Facebook at its word? The social-media platform’s record isn’t exactly one that inspires trust.
Luis Miguel is a marketer and writer whose journalistic endeavors shed light on the Deep State, the immigration crisis, and the enemies of freedom. Follow his exploits on Facebook, Twitter, Bitchute, and at luisantoniomiguel.com.