After a deepfake voice fooled her grandfather, this founder built on-device detection

When the call came, Tarini Padmanabhuni's grandfather believed he was speaking to his brother. The voice said his brother had been kidnapped and that a ransom was the only way to get him back. He paid, and only later learned that his brother had been somewhere else entirely. The voice was a deepfake. "What stayed with me wasn't the money," Padmanabhuni says. "It was that he had no way of telling."

That was about two years ago. DetectifAI, the San Francisco company she founded, now aims to keep others from being fooled the same way.

A growing problem, and a market to match

According to the FBI, Americans lost close to $900 million to AI-driven scams last year, up 24% from 2024. People 60 and older lost twice as much as those aged 50 to 59.

Competition is not scarce: Reality Defender, Pindrop, Resemble AI, Microsoft Azure AI Content Safety and Nuance. But existing products run in the cloud, Padmanabhuni argues, so phone makers cannot build them into their devices.

Detection on the device instead of the cloud

Rather than shrinking large cloud models to fit a phone, DetectifAI designs compact models from the start, small enough to run inside a phone's operating system. The goal is an instant verdict on whether a voice is AI-generated during calls and in voice messages, without the audio ever leaving the device.

The company sells first to phone manufacturers, licensing its tools so detection ships as a built-in feature of the system. Padmanabhuni compares it to AT&T's role in the original iPhone launch: the first phone maker to ship DetectifAI gains an edge, and deepfake detection becomes a standard spec, like camera resolution.

An SDK, early revenue and a small seed round

The core of the product is DetectifAI's software development kit, code that other companies can build into their products. A secondary revenue stream will come from licensing the technology to businesses and fraud-prevention firms.

The startup already has early revenue and handles more than 100,000 calls a month for financial institutions in India, placed by AI voice agents handling debt collections and loan documents, with deepfake detection and speaker verification on every call. Padmanabhuni declined to name customers.

DetectifAI has raised a small seed amount from Josh Constine, formerly a TechCrunch editor, and Manohar Kamath, and is among the startups vetted to compete in Startup Battlefield at TechCrunch Disrupt, October 13 to 15 in San Francisco.

Asked about the most rewarding moment so far, she points to a small WhatsApp beta in which users forward suspicious voice notes and get back an assessment of whether they are real. One tester said they would pay for it without hesitation. "My grandfather didn't have that," she says.