Phixo: AI Phishing Extension That Explains Red Flags

Most phishing filters just throw a "this is suspicious" warning and call it a day. No context, no reasoning—just a binary flag that leaves users confused. That's the problem I wanted to solve with Phixo, a Chrome extension I built as a solo developer.

I started looking into how ordinary people interact with phishing warnings. My mom, my neighbors, even some coworkers—they all click through warnings because they don't understand why the email is flagged. So I built something that breaks down the evidence: here's the header inconsistency, here's the spoofed domain, here's the urgency language pattern. It's like having a security analyst explain it in plain English.

Phixo works inside Gmail and Outlook web clients. When you open an email, it scans the content in real time—checking sender reputation, analyzing URL redirect chains, spotting mismatched display names, and matching known social engineering templates. If it finds something sketchy, it blocks the link at the browser level and lays out each suspicious element in a small panel above the email. The goal is not just protection but education—users gradually learn to spot patterns themselves.

Right now it's still early. I'm actively improving the detection model and the explanation quality. The hardest part is balancing false positives: explain too many benign emails and users get desensitized. I'm betting on precision over volume—fewer alerts, each one well-justified.

Technical side: it runs a lightweight on-device model for quick checks (domain age, known bad TLDs, common phishing