Stop scams before people click.
WhatsGuard is an open-source, local AI scam-warning tool for WhatsApp. It helps people spot risky messages, phishing links, impersonation attempts, and urgent payment scams while keeping private conversations on their own PC or Mac.
Possible phishing link, urgent tone, unverified sender. Pause and verify through the official channel.
your chats
The problem
Scams today aren't just convincing — they're custom-built for you.
They look like bank alerts, delivery updates, family emergencies, job offers, investment groups, school payments, and account warnings. WhatsGuard is designed to create a calm pause at the moment of risk.
Scammers create urgency
“Verify now.” “Pay now.” “Reply now.” The goal is to stop people thinking clearly.
Families need a second pair of eyes
Many users are not only worried about themselves. They are worried about parents, children, staff, and community members.
Privacy cannot be the price of safety
Fraud protection should not require people to casually upload private conversations to someone else’s server.
Human first
This is for people who worry about other people.
The person installing WhatsGuard may be a daughter protecting her mother, a parent guiding a child, a teacher running a safety session, or a community worker supporting seniors.
Choose your role
Different people can help in different ways.
WhatsGuard is not only an app. It is a community defense project for people who share the same anti-scam and online-safety mission.
Protect yourself and people you care about.
Use the free version to get warnings before clicking suspicious links, replying to impersonation messages, or sending money under pressure.
- Install once on PC or Mac
- Warnings in plain language
- Built for everyday users, not IT teams
Turn scam awareness into practical protection.
Use WhatsGuard in workshops, scam clinics, parent sessions, senior programs, and digital-literacy campaigns.
- Scam-message clinic format
- Community education materials
- Feedback loop for vulnerable groups
Help inspect, improve, and harden the tool.
Contribute to the local AI runtime, model evaluation, security review, prompt-injection testing, accessibility, and localization.
- Open-source roadmap
- Security and privacy review welcome
- Translation and UX contributions
Support public-interest scam prevention.
Partner with us to test the free version with seniors, families, students, volunteers, and frontline community helpers.
- Small pilot groups
- Localized education kits
- Measurable feedback and public learning
Privacy model
Your messages should stay with you.
WhatsGuard is designed around local analysis. The aim is to help people detect high-risk messages without asking them to surrender private chats to an outside reviewer.
Message appears in your WhatsApp session
WhatsGuard watches for risky patterns such as phishing links, urgent payment requests, and suspicious identity claims.
The safety check happens on your computer
The AI engine runs locally on your own device, helping produce a warning and explanation without cloud scanning by default.
You get a calm warning
The user sees why the message may be risky and is encouraged to pause, verify through official channels, or ask someone trusted.
AI in defense mode
Use AI to help ordinary people detect manipulation, urgency, suspicious links, and impersonation patterns.
Runs on your computer
The detection engine is designed to run on the user’s own PC or Mac, reducing unnecessary exposure of private conversations.
Open source
Trust should be inspectable. Developers, privacy reviewers, and community groups can examine and improve the project.
Made for vulnerable groups
The experience is designed for families, seniors, kids, caregivers, and non-technical users who need simple warnings.
Community pilot
Help build a community shield against scams.
We are not looking for “beta testers” in the usual software sense. We are inviting early community protectors: people and organizations who want to help families pause before risky clicks, replies, and payments.
Plain answers
Questions people ask first.
Is this an official WhatsApp product?
No. WhatsGuard is an independent protection project. It is designed to help users identify suspicious messages, but it is not made by or endorsed by WhatsApp.
Will WhatsGuard see or upload my chats?
The project is designed around local analysis on your own computer. The goal is to provide scam warnings without requiring users to upload private conversations for outside review.
Can it detect every scam?
No. No tool can guarantee that. WhatsGuard is a warning tool that helps people pause, check, and verify before acting.
Why open source?
Because privacy and security claims should be inspectable. Open source lets developers, researchers, and community partners review how the tool works.
Who should use the free version?
Families, seniors, parents, community helpers, online-safety advocates, NGOs, developers, and researchers who want to help build practical scam defense.
What feedback do you need?
Installation friction, unclear warnings, false alarms, missed scams, privacy concerns, translation issues, accessibility problems, and ideas for vulnerable users.
Get involved