This Starts With a Scam Call
Two years ago, Patrick Coughlin's mother got a phone call. On the other end was her daughter — crying, saying she'd been kidnapped, demanding ransom.
It was an AI voice clone. A few seconds of audio is all it takes to generate a synthetic voice indistinguishable from the real thing.
Patrick was a senior vice president of security products at Cisco at the time. He'd spent his career in cybersecurity and defense. And his mother still got taken.
That became the reason he started a company.
July 7: Savi Security Goes Public
Patrick and his brother Ryan built Savi Security. Ryan had done consumer products at Apple and Spotify.
On July 7, they officially launched a consumer-facing AI anti-scam app, announcing a $7 million seed round led by Acrew Capital.
The app covers iOS and Android, and it does one thing directly: detect AI-scam signatures across calls, texts, and emails in real time, warning users before they get taken.
Why This Is a Real Pain Point
FBI data: in 2025, AI-related scams cost the U.S. over $10 billion.
And it's climbing. AI voice cloning, deepfake video, hyper-personalized phishing — traditional anti-fraud can't keep up.
A few seconds of audio clones a voice. The you your parents hear begging for help on the phone may not be you at all.
What's Interesting Here
Most AI security products are built for enterprises — firewalls, threat detection, compliance audits. Savi picked a different lane: it's for regular people.
Wrapping professional-grade security capability into something ordinary users can actually use. Totally different logic from enterprise security.
The founding team fits — Patrick brings cybersecurity and defense chops, Ryan brings consumer product sense. One knows the tech, one knows how to turn tech into something people can use.
But Let's Cool the Hype
$7 million seed is very early. How well does the app actually detect scams? What's the false-positive rate? Can it keep up as scammers upgrade their own tooling? None of this is proven at scale yet.
And AI-scam vs. anti-scam is fundamentally an arms race. Scammers use stronger generative models too. The signatures you catch today get bypassed tomorrow. Whether Savi can keep running ahead of the criminals — that's the real test.
Then there's the old consumer-security problem: will people keep paying for it? Antivirus already proved this once — free options ate most of the paid market. Why would Savi be different?
That said, the founding motive is real. Patrick's mother's story isn't marketing — it actually happened. New risks from AI need new defensive tools. Whether this direction holds and how far it goes depends on whether the product can truly hold up. But at least someone's seriously working on it.