Facebook Offers to Pay For Users’ Private Data Including Voice Records

After facing criticism from the public and lawmakers over the non-consensual collection of data, Facebook has started offering to pay its users for personal information including recordings of their own voice.
Sputnik

The recordings, made through its new market research app Viewpoints, will help to train the speech recognition system that powers Facebook’s Portal device which the company has high hopes will become a contender with other voice assistants, including Amazon’s Echo speakers and its Alexa virtual assistant.

The Viewpoints app, which was revealed three months ago as a portal for conducting surveys about user experiences, began to invite users in the US to say “Hey Portal” and the names of up to 10 friends. The recording process is giving users points which can be converted into a $5 cash reward.

Facebook has indicated that Viewpoint was initially designed to “improve speech understanding” by training machine-learning algorithms, which require a large number of examples to improve their accuracy and performance.

“Participants record phrases within the app, which helps us improve name pronunciation recognition in our products to better serve the people that use them,” a Facebook spokesperson said.

In the description on the service’s website, the social media giant wrote that the information gathered with Viewpoint activity won’t be shared on Facebook or on other linked accounts without users’ permission, or be sold to third parties. However, the Viewpoints data policy notes that some information collected using the app, such as payment and device data, can be used to personalize other Facebook apps and target advertising. It could also be shared with “research partners,” including academics, publishers and advertisers, however, Facebook promised that users will be notified if this happens.

Makers of smart speakers including Facebook, Amazon, Apple and Google faced criticism last year when it was revealed that they were sending users’ voice recordings to human moderators without the consent or awareness of the customers. The media companies promised to stop such practices and ended their partnerships with several data centres worldwide.

Discuss