DetectifAI develops on-device deepfake voice detection for smartphones

Origin and Motivation

Tarini Padmanabhuni founded DetectifAI after her grandfather was defrauded by a deepfake voice impersonating his brother. The incident involved a ransom payment for a fabricated kidnapping, highlighting the difficulty in distinguishing AI-generated voices from real ones.

The FBI reported that Americans lost close to $900 million to AI-driven scams last year, a 24% increase from 2024. Individuals aged 60 and older experienced twice the losses compared to those aged 50 to 59.

Addressing Current Limitations

Existing deepfake voice detection solutions from companies like Reality Defender, Pindrop, and Microsoft Azure AI Content Safety typically operate in the cloud. This approach prevents direct integration into mobile devices, leaving users vulnerable during real-time interactions.

DetectifAI's approach differs by designing AI models that are compact enough to run directly within a smartphone's operating system. This eliminates the need for audio data to leave the device for processing.

Technology and Implementation

The core of DetectifAI's offering is its software development kit (SDK), which allows other companies to integrate the detection capabilities into their products. The technology provides instant verdicts on whether a voice is AI-generated during calls, in voice messages, and other audio formats.

DetectifAI plans to initially sell its technology to phone manufacturers, licensing its software tools for built-in deepfake detection. This strategy aims to establish deepfake detection as a standard feature in smartphones, similar to camera resolution.

Revenue Streams

The primary revenue stream for DetectifAI will come from licensing its SDK to phone manufacturers. A secondary revenue stream is planned through licensing the technology to businesses and fraud prevention entities.