AI Voice Cloning: A Growing Threat to Personal Privacy and Ethics in Technology

In a recent high-profile case, a physical education teacher and former athletic director of a Baltimore County high school was arrested for using an AI voice cloning service to frame the school principal. However, interest in AI-powered voice cloning technology has grown as services become better at sounding more human. Cases of improper, if not criminal, use of apps and software for voice cloning are also on the rise.

The Baltimore case involved a recording that circulated on social media in January, allegedly featuring fake audio of Pikesville High School Principal Eric Eiswert making racist and anti-Semitic comments. Experts noted that the recording had a “flat tone,” unusually clean background sounds, and lacked coherent breathing sounds or pauses.

Baltimore County police traced the recording back to Dazhon Darien, a former athletic director at the school who used the school’s computers to access OpenAI tools and Microsoft Bing Chat services. He was also linked to the release of the audio via an email address and associated recovery phone number. It is unclear which AI voice platform Darien used.

Darien was arrested at the airport on charges including theft (for investigating potential mismanagement of school funds), disrupting the operations of a school, retaliation against a witness, and stalking. Other notable cases have also occurred in Pakistan, where Imran Khan’s political party used ElevenLabs to replicate his voice during an electoral campaign. In Texas, two companies were linked to a fake robocall pretending to be President Joe Biden telling people not to vote. The Federal Communications Commission banned AI robocalls in February. Additionally, OpenAI decided in March to hold back its AI text-to-speech generation platform for public use due to concerns about guardrails around the technology. US lawmakers have introduced bills such as the No Fakes Act and No AI Fraud Act that seek to prevent tech companies from using an individual’s face, voice or name without their permission.