After Google, Apple, and Amazon publicly admitting that they have been using third-party contractors and employees to listen to snippets of recordings from their devices, Facebook joins the list after the tech giant admitted that its employees are listening to users’ voice messages.
People who are familiar with Facebook’s operations — who requested for the protection of their anonymity in fears of losing their jobs in the company — reveals that Facebook is using real people to transcribe audio messages sent by users through Facebook’s messaging application, Messenger.
The practice has alarmed contractors saying that Facebook gave them voice recording to be transcribed. Facebook failed to advice where the recordings were taken from and how were they collected, and people familiar to the matter also said that they are not sure what’s the need for transcription, in the first place.
Contractors tasked to do the human review of voice recording transcription said that they are hearing Facebook voice chat recordings and they sometimes include vulgar and sensitive conversations.
Facebook did not deny the allegation of practicing human review of recording sent through their platform. The social media giant clarified that this activity had been stopped, following the intensified scrutiny on the practice involving other tech giants like Apple and Google.
“Much like Apple and Google, we paused human review of audio more than a week ago,” the company said Tuesday.
In their defense, Facebook said that only those users who enabled the voice-to-text transcription feature in Messenger were affected by the practice, as they are aiming to check whether the artificial intelligence (AI) that is used to do the transcriptions are accurate. The San Francisco-based social media platform also clarified that the recordings were non-identifiable and anonymized.
Nonetheless, authorities have launched an investigation on the possible violations of user privacy in relation to the human review practice of Facebook.
On Wednesday, the Irish Data Protection Commission, the agency that is tasked to oversee Facebook in Europe, launched an investigation against Facebook’s practice to determine whether EU residents were affected, and if, their privacy was compromised and violated. The agency said that Facebook could have been violating the EU’s strict privacy laws.
The Irish Data Protection Commission said it is “seeking detailed information from Facebook on the processing in question.” The Irish regulatory body has also looked into similar incidents involving tech companies like Google, Amazon, and Apple.
“We are now seeking detailed information from Facebook on the processing in question, and how Facebook believes that such processing of data is compliant with their” obligations under the EU’s privacy rules, the Irish regulator said.
Earlier this month, smartphone giant, Apple, also admitted in having the practice of contracting the service of several people to listen to clips of recordings made by iPhone and other Apple devices users through Siri to “grade” how the AI responds to people’s inquiries through the service.
However, it was revealed that Apple has also been collecting accidental conversations. This happens when Siri is activated when it mistakenly hears the “wake word,” “Hey Siri.” Interestingly, this happens a lot of times.
After the disclosure of Apple’s practice, the company said that while they are investigating the practice, they are suspending its operations involving Siri recordings.
“We are committed to delivering a great Siri experience while protecting user privacy. While we conduct a thorough review, we are suspending Siri grading globally. Additionally, as part of a future software update, users will have the ability to choose to participate in grading,” the company said in a statement.
Similarly, Google was slammed when it was found out that the company is also using humans to listen to recordings for their smart speaker devices. The company said that linguistics experts are listening to “snippets” of recordings made by users to improve the device’s artificial intelligence for voice recognition technology used in their smart home systems and smartphones. Nonetheless, they clarified that the tapes were randomly snipped and are unnamed.
“We partner with language experts around the world to improve speech technology by transcribing a small set of queries – this work is critical to developing technology that powers products like the Google Assistant,” Google said. “Language experts only review around 0.2% of all audio snippets, and these snippets are not associated with user accounts as part of the review process.”