WhatsApp Tests New AI-Powered Scam Alert Feature to Detect Fraudulent Messages

WhatsApp is testing a new AI-powered security feature that automatically identifies fraudulent messages on users' devices. The system analyzes suspicious message patterns locally on the smartphone without transmitting data to external servers, protecting user privacy. When a message is flagged as potentially dangerous, users receive a warning and can choose to block the sender or report the message.

New Scam Alert Feature for WhatsApp

WhatsApp is developing a new security feature designed to protect users from fraud attempts using artificial intelligence. The feature, currently in testing phase, aims to automatically identify suspicious message patterns and provide warnings directly within the application.

How the Technology Works

The key aspect of this new Scam Alert feature is that analysis occurs directly on the user's device. According to WhatsApp, incoming messages are analyzed locally, meaning data remains on the user's smartphone and is not transmitted to external servers or accessed by Meta. The AI models search for specific patterns typical of fraudulent activities on the device itself. When a message is flagged as potentially dangerous, a warning is generated. This approach responds to the increasing complexity of fraud schemes in the digital space while maintaining end-to-end encryption and data privacy.

User Options When Alerts Appear

When a message is flagged, users have several options available directly from the warning dialog. They can block the sender immediately, report the message to the platform operator, or choose to ignore the warning and continue the conversation if they believe the contact is safe. This integration into the chat history aims to reduce response time and lower the barrier to taking security measures.

The Scale of Fraud Losses

The importance of such protective mechanisms is underscored by current data from the U.S. Federal Trade Commission (FTC). In 2025, losses from fraud on social media platforms worldwide totaled 2.1 billion US dollars. Messaging service users are particularly affected, with WhatsApp users alone experiencing losses of 425 million US dollars.

Implementation Timeline

With the introduction of the AI-powered Scam Alert, Meta is attempting to reduce these damage sums and strengthen confidence in the platform's security. However, the timeline for when this feature will exit beta status and become available to all users worldwide has not yet been specified.