DetectifAI develops voice-authentication to combat deepfake scams

San Francisco-based startup DetectifAI is developing voice-authentication models designed to identify synthetic, cloned, or manipulated speech. The company aims to combat fraud where criminals use AI-generated voices to impersonate relatives, businesses, or financial institutions.

Founder Tarini Sai Padmanabhuni established the company after her grandfather fell victim to a ransom scam involving an AI-generated imitation of his brother’s voice. To address the growing threat, DetectifAI is designing compact AI models capable of running directly on a smartphone’s operating system or within telephony edge systems. This approach allows for real-time detection of deepfakes during calls or in voice messages without the audio needing to leave the device.

The company is targeting phone manufacturers, financial institutions, and contact centers. DetectifAI currently processes over 100,000 calls per month for financial institutions in India to assist with deepfake detection and speaker verification. This development comes as the FBI reported nearly $900 million in losses to AI-driven scams in the United States, a significant increase from previous years.