Artificial Intelligence to Detect Scammers in WhatsApp Conversations

“Russia Today” – WhatsApp announced a new security feature designed to help users verify the identity of people they communicate with, especially if those individuals are not saved in their contact list.

The new tool, called “Scam Alert,” relies on artificial intelligence to analyze messages received from unknown numbers. It examines “conversation structure and linguistic cues” to assess the likelihood that a message is part of a scam.

When the system suspects fraudulent intent, it displays a warning within the chat that is hidden from the other party. Users are free to block the sender, report the message, or continue the conversation as usual.

Users must manually activate the feature through the settings. Once enabled, the app downloads a machine-learning model that runs locally on the user’s device, meaning that Meta, the parent company of WhatsApp, does not receive any copy of the conversation content nor share it with any third party, according to the company’s assurances.

The scam-alert system was built using previously reported conversations that users flagged as scam attempts, which were used to train the AI model to recognize suspicious messages.

The feature is currently being rolled out in a limited beta test to a select group of early users. WhatsApp plans to conduct extensive testing of the system and has invited researchers to report any vulnerabilities before a wider release to all users in a later phase.