What eventually becomes cutting-edge in retail and payments often originates in places that are dedicated to security, not commerce. New technology often starts out at the U.S. Department of Defense (DOD) or NATO, for instance. Former Israeli intelligence agents or military members have taken the knowledge they gained during their service and contributed to artificial intelligence (AI), fraud prevention and other efforts that do make -- or can end up making -- a difference in the daily lives of consumers. Now, the time might be right to at least consider adding prisons to that mix. Let's not get ahead of ourselves -- there is little chance that any prison system will become a miniature Silicon Valley.

Prisons across the United States are reportedly building biometric databases that include voice recordings of incarcerated people, according to The Intercept. The report cites contracting documents for the state of New York's prison system, as well as statements from officials in Texas, Florida, Arkansas and Arizona confirming that prisons are actively using voice recognition technology that can extract and digitize voices to create unique and identifiable biometric signatures known as voice prints. According to the prisons, the systems are being used to improve security in the facilities and to combat fraud. Prisons can use the biometric identifier to track phone calls and find past communications that match the voice print of a particular person. The systems can also flag calls deemed to be suspicious to allow investigators to review the communications.

Voice recognition has come on in leaps and bounds over recent years and a new AI can now pick out an individual's voice from a crowd. The system uses machine learning to identify the voice prints of a group of speakers, then reconstructs what each person has said. This overcomes what researchers have dubbed the'cocktail party effect' and could lead to improved smart assistants and better automated transcription of speech. Voice recognition has come on in leaps and bounds over recent years and a new AI can now pick out an individual's voice from a crowd. In tests the system could differentiate the voices of up to five people at the same time.

For all of Alexa's many skills, one major thing Amazon's voice assistant is currently missing is the ability to distinguish who is actually speaking to it. According to a new report from Time magazine, however, the online retail giant is currently working to make that a reality. As Time's sources inside Amazon explained, the feature would use a "voice print" to verify each speaker's identity and primary account holders could set Alexa to require their voice print for specific commands. That means that the person with the credit card tied to the account could set their Echo to require their voice for purchases and restocking the fridge. The Voice ID feature, which has reportedly been in development since 2015, could also help ease some of those Alexa headaches and accidental shopping sprees.

Amazon is working to make its Alexa a better listener. The Seattle-based technology giant has been developing a feature that would allow the voice assistant that powers its Echo line of speakers to distinguish between individual users based on their voices, according to people familiar with Amazon's Alexa strategy. The sources declined to be identified by name because they are not authorized to talk about the company's future product plans. An Amazon spokesperson declined to comment. Alexa, like Apple's Siri or Microsoft's Cortana, can interpret and respond to voice commands such as "How's the weather?" or "What movies are playing tonight?"