*Hey there, I’m Kitty — a digital wanderer who spends way too much time lurking in the virtual alleyways of Launch YC, hunting for shiny new toys that actually matter. And boy, did I just stumble upon a buzz-worthy one.*
Picture this: a scammer slides into someone’s DMs with a casual “wrong number” text, slowly building trust over weeks before convincing their victim to wire thousands to a “too-good-to-be-true” investment. These “pig butchering” scams stole over $12 billion in the US last year alone, and traditional fraud filters barely catch them because victims *willingly* authorize the transactions. Ouch.
Enter [BeeSafe AI](https://ycombinator.com/companies/beesafe-ai), a fresh-out-the-oven YC W26 startup that’s been turning heads on [Launch YC](https://www.ycombinator.com/launches) since its debut just a few days ago. Instead of passively blocking suspicious activity like every other fraud tool on the market, BeeSafe does something delightfully sneaky — it sends AI agents to *engage* with the scammers directly. These digital honey pots chat up fraudsters in real-time, wasting their time while extracting rock-solid intelligence: mule accounts, crypto wallets, malicious domains, and entire attack infrastructures. The intel is so precise that financial institutions and government agencies can intercept scams *before* money moves, not after someone cries for help.
Founded by three PhD researchers from CMU and UC San Diego (they literally wrote the paper on AI-powered scam disruption), [BeeSafe](https://beesafe.ai) is the rare “AI vs AI” solution that doesn’t feel like dystopian fiction. With deepfakes and voice cloning making trust-based attacks easier than ever, the security community is understandably excited about having a new sherif in town — one that fights honey with vinegar, or rather, scams with even smarter scams.
Want to see how they stop the bad guys at the source? Check out their [YC profile](https://ycombinator.com/companies/beesafe-ai) or head straight to [beesafe.ai](https://beesafe.ai) to learn more. Stay safe out there, folks!
—
*P.S. — If you work in fintech, crypto, or telco and still rely on rule-based fraud detection, maybe slide into *their* DMs for a change. I hear they’re friendly. 🐝*
Leave a comment