Samsung on Tuesday clarified the voice-recognition capabilities of its smart TVs and said its sets do not monitor living room conversations.

The company has also updated its privacy policy to more fully explain how its voice recognition works.

Yesterday, word circulated that Samsung's smart TV privacy policy warned users that spoken words containing personal or other sensitive information "will be among the data captured and transmitted to a third party."

Those third parties, according to Samsung's policies, perform tasks like converting speech to text, while data collection helps the company improve its services. The feature can be turned off via the settings menu, but may prevent the user from accessing all voice-recognition functions.

Amidst concern that Samsung was eavesdropping on people's conversations, the company issued a blog post titled "Samsung Smart TVs Do Not Monitor Living Room Conversations."

Samsung offered an explanation for how its voice-recognition software works: a mic embedded in the TV responds to pre-determined TV commands, like channel or volume changes, and that data is not stored or sent anywhere; a second microphone, however, inside the remote must interact with a server in order to search for and serve up content. "This interaction works like most any other voice-recognition service available on other products including smartphones and tablets," Samsung said.

The uproar prompted Samsung to change it privacy policy to further explain voice recognition: "To provide you the Voice Recognition feature, some interactive voice commands may be transmitted (along with information about your device, including device identifiers) to a third-party service provider (currently, Nuance Communications, Inc.)," the clarification said.

Nuance then converts the commands to text, "and to the extent necessary to provide the Voice Recognition features to you," Samsung said, adding that the company may collect—and your device may capture—voice commands and associated texts "so that we can provide you with Voice Recognition features and evaluate and improve the features."

PC Magazine