Voiceprints raise surveillance fears
23 January 2018 14:24 GMT

Recently uncovered evidence of how the NSA developed a voice-print monitoring system could impact on perceptions of the technology.

In an Intercept piece on Friday, reporter Ava Kofman detailed the secret history of the NSA’s speaker recognition systems, dating back as far as 2004.

One of the programs was a system known as Voice RT, which was able to match speakers to a given voiceprint (essentially solving the Waldo problem), along with generating basic transcriptions.

According to classified documents, the system was deployed in 2009 to track the Pakistani army’s chief of staff, although officials expressed concern that there were too few voice clips to build a viable model. The same systems scanned voice traffic to more than 100 Iranian delegates’ phones when President Mahmoud Ahmadinejad visited New York City in 2007.

The documents also describe how the NSA used the technology during Operation Iraqi Freedom to match an audio recording to former leader Saddam Hussein’s “voiceprint.” The NSA also used the technology to compare the voice of a captured suspect with previous audio recordings from terrorist Abu Hakim to confirm that the suspect was not a match.

In order to test their technology, analysts at the NSA compared old intercepts and audio files relating to Ron Pelton, a former NSA analyst who sold details about several secret US programs to the Soviet Union. At the time, the agency failed to identify Pelton through human voice identification. However, in 2006, the agency was able to automatically match Pelton’s voice using the technology.

“Had such technologies been available twenty years ago, early detection and apprehension could have been possible, reducing the considerable damage Pelton did to national security,” the document states.

It remains to be seen if the revelations will impact the public's view of how IOT devices like Alexa and Google Home collect and use voice data.

Albert Gidari, director of privacy at the Stanford Center for Internet and Society, told The Verge that the government may have a right to compel firms to divulge voice prints. “To the extent platforms store biometrics, they are vulnerable to government demands for access and disclosure,” says Gidari. “I think the government could obtain a technical assistance order to facilitate the scan, and under [the technical assistance provision in] FISA, perhaps to build the tool, too.”