Fraud Zer0

Fraud Zer0 detecting fake email on outlook

Inspiration

Every day, people—especially students, seniors, and first-time internet users—lose money to scams that look too real to be fake. From fake job offers and phishing emails to fraudulent loan sites, millions are tricked before they even realize what happened.

Even today, most tools aren’t 100% accurate at blocking these frauds or teaching users how to stay safe. We wanted to change that. Fraud Zer0 was born to not only protect people from scams but also teach them how to recognize and avoid them, empowering underserved users who often lack access to reliable digital financial protections.

What it does

Fraud Zer0 is a Chrome extension that acts as your personal financial firewall. It detects and blocks suspicious online activity such as:

-Fake job offer emails

-Phishing pages that imitate real banks

-Predatory loan sites

-Fake links in emails or webpages

When a risk is detected after scanning, the extension immediately displays a small banner explaining why the page, email, or link looks suspicious. It can also provide a short, friendly lesson on how to identify and avoid similar scams in the future, helping users get smarter and safer every time they browse.

How we built it

We built Fraud Zer0 using:

-Claude API to analyze messages for potential scam indicators.

-Fetch AI to provide foundational training data such as examples of fake URLs, suspicious email domains, and fraudulent messaging patterns.

-Eleven labs to initiate voice alerts

-OpenAI Vision to capture and interpret on-screen content, sending snapshots for analysis to our AI models (Claude and Gemini).

-Chrome Extension Manifest V3 for the frontend and content scripts, ensuring modern, secure browser integration.

-Gemini AI to analyze web pages and links for scam patterns and generate clear, educational explanations for users.

Challenges we ran into

Creating real-time scam detection that works across websites like Outlook and Gmail was tough — especially since email clients use complex, dynamic HTML. We spent hours optimizing content scripts to correctly parse and analyze messages without breaking page functionality.

Setting up and synchronizing multiple AI APIs like Gemini and Claude was another major challenge. The initialization process, authentication handling, and ensuring smooth communication between them took a lot of time and debugging.

Balancing protection and learning was also difficult — we didn’t want the extension to just block everything, but to explain why something was risky in a clear, human tone.

Accomplishments that we're proud of

-Built a working demo that flags financial scam emails (fake job offers, prize/ reward scams) in Outlook Web.

-Implemented a pop-up system that intercepts suspicious links and prevents navigation to fraudulent sites.

-Created an educational feedback loop that teaches users why something was flagged, not just warns them.

Designed an accessible interface that anyone—regardless of tech skill—can use

What we learned

We learned how to leverage AI to make scam detection smarter while simultaneously teaching users how to stay safe.

We also learned that underserved users often lack access to reliable tools for digital financial safety, and that education plus real-time protection can help them navigate the online financial world more confidently.