A person’s voice can allow machine learning to guess a person’s age, gender, ethnicity, socioeconomic status and more, Wired reports. The voice recognition industry has grown into a $20 billion business. Users’ so-called “voiceprints” could be used for targeted advertising or the location where they are stored could be used by hackers to impersonate the user, such as robocall scams where criminals have a recording of a user saying “yes.” Privacy researchers are investigating methods such as anonymization or obfuscation to protect user voice data.
User 'voiceprints' may pose privacy challenge
Related stories
Notes from the IAPP Europe: It's a GDPR week
Notes from the Asia-Pacific region: More new problems to solve as AI, data march forward
Proposed moratorium on state-level AI regulation heads to US Senate
'We cannot be left behind:' How Canada is balancing AI regulation, innovation
The ethical use of AI in advertising