OmniSpeech targets voice deepfake detection market with Zoom integration

Maryland-based OmniSpeech has released its audio deepfake detection software to the Zoom Marketplace to secure video calls against AI-powered fraud. Now, it is seeking additional funding to scale its product development.

The company’s AI Detect uses proprietary algorithms to detect and deliver alerts of any voice deepfakes with its zero-shot technology.

User implementing OmniChannel get a “Start Scan” button on the Zoom interface. Pressing it activates the software, which assigns a color to each voice it detects. Red means stop – it’s a deepfake.  Green means the voice is likely human, and yellow indicates uncertainty, and therefore a need for caution.

“Zero-shot” refers to the software’s ability to recognize fakes generated by sources it has not been trained on.

OmniSpeech says in its announcement that the software analyzes audio from any source, which means that Zoom can also be used as a universal playback and detection environment.

“OmniSpeech AI Detect is an example of how partners are extending the power of Zoom’s open platform to tackle everyday challenges,” says Zoom Chief Ecosystem Officer Brendan Ittelson in the company announcement. “AI deepfake detection in Zoom Meetings using Zoom RTMS helps individuals and organizations preserve trust, validate authenticity, and protect their conversations in real time — all without changing how they meet.”

The AI Detect API is available now, and the company is marketing it for a wide range of applications, from HR use for remote interviews and onboarding, compliance and security against fraud and social engineering and media and content reviews to everyday authenticity checks by consumers.

OmniSpeech was one of three organizations splitting $35,000 in prize money from the FTC in 2024 as co-winners in the smaller organizations category of its Voice Cloning Challenge.

In order to expand its product development capabilities, the company is now looking to raise seed funding early in 2026, according to a report by TECHNICAL.LY.

Article Topics

biometrics  |  deepfake detection  |  deepfakes  |  funding  |  OmniSpeech  |  spoof detection  |  Zoom