The contractors said that they were paid to listen in on countless, occasional vulgar conversations among Facebook Messenger users, though the company never explained to them why it needed the transcriptions. The social media giant tells the outlet that the only users who are affected are those who chose the option to have their audio chats transcribed.
The sources state that the social networking giant has hired outside contractors to go over the audio clips, with every single word (good and bad, secret and explicit conversations) being transcribed. The Irish regulator, which takes the lead in overseeing Facebook in Europe, said it is already looking at similar data processing by Google, Apple Inc. and Microsoft Corp.
Facebook's lead supervisory authority on data protection within the European Union is Ireland's Data Protection Commissioner.
In Facebook's case, the company acknowledged Tuesday that it paused the practice earlier this month after those other companies started to attract scrutiny over basically the same thing.
Facebook did not respond to a request for comment from AFP. According to the company, Facebook "paused human review of audio more than a week ago", following suit with Apple.
The Facebook data-use policy, revised previous year to make it more understandable for the public, includes no mention of audio.
According to Bloomberg, Facebook first started allowing users of their Messenger service to have their recorded audio conversations transcribed in 2015.
It added this had been done only when users had opted in to transcription services and given permission for microphone access.
Just recently, Facebook and the U.S. Federal Trade Commission reached a settlement regarding some shady privacy practices, and now we learn that there is much more to the story. Only that "systems automatically process content and communications you and others provide" with "vendors and service providers who support our business" but not mentioning the possibility of people transcribing personal messages.
Facebook's help center simply mentioned in a note that the Voice-to-Text feature uses machine learning; there is no mention of the fact that actual people may be listening to the voice recordings.