The use of voice recognition is becoming prominent and is on the rise in many pieces of software. Yet there are vulnerabilities and shortcomings of algorithms and technologies that allow attackers to perform voice morphing, theft of voice histories, etc., to perform other destructive attacks and influence companies, individuals, politics, legal proceedings, and more. The possibilities are limitless. This talk identifies those vulnerabilities, risks, and preventions.