Meta is launching an optional WhatsApp scam detection feature that uses on-device machine learning to flag suspicious messages. According to a report by The Verge, this cybersecurity update is currently rolling out in a limited beta. The tool aims to protect users from deceptive messaging and financial fraud directly on their mobile devices.
How WhatsApp scam detection Works
The new system operates directly on the user’s smartphone to identify potential threats. If the machine learning model flags a conversation as a potential fraud attempt, a warning appears within the chat interface. Notably, this warning remains completely invisible to the sender of the message. Users can then choose to block the contact, report the conversation, or continue chatting if they believe the warning is incorrect.
If a user decides that a warning is a false positive, they can mark the chat as trusted. Once marked, the warning disappears, and the WhatsApp scam detection system will not flag that specific conversation again. Additionally, users who trust a chat can opt to share the last five received messages with the platform to help train and improve the accuracy of the detection model.
User Control and Privacy Protection
Privacy remains a central component of this new security implementation. Meta stated that no message content leaves the user’s device for classification purposes when the feature is active. Furthermore, nothing is automatically reported to WhatsApp, Meta, or any third party. Users retain full control over the security tool and can choose to turn off the detection feature at any time through their settings.
The Growing Threat of Digital Fraud
The introduction of this tool comes as digital messaging platforms face increasing security challenges. With more than 3 billion users globally, the apps and messaging application has become a primary target for various fraudulent schemes, including wire transfers and complex financial scams. Data from the Federal Trade Commission (FTC) indicates that victims reported losing 425 million dollars to scams on this platform alone, out of 2.1 billion dollars lost to social media fraud in 2025.
Future Outlook for Platform Security
This update represents a continuation of Meta’s efforts to secure its messaging ecosystem. Earlier this year, the company introduced a similar security feature designed to detect fraudulent activity during device linking requests. As the limited beta progresses, the company plans to refine the on-device machine learning models to better identify evolving deceptive tactics while maintaining strict user privacy standards.
