Instant messaging application WhatsApp has announced the introduction of a new scam alert feature to protect users from online fraud. This feature will use artificial intelligence (AI) to review messages that are not received from people in the user’s contacts.

Through the new tool, WhatsApp will analyze the conversation style and linguistic cues of messages to assess whether the relevant message is part of a potential scam or fraud.

If WhatsApp suspects that a conversation is fraudulent, a warning will be shown to the user within the chat. This warning will only be visible to the relevant user and will not be known to the other person.

After receiving the warning, the user will have the option to block or report the relevant account, while he can also continue the conversation if he wants.

According to WhatsApp, this feature must be enabled by the user himself. When enabled, the app will download a machine learning model that will analyze the messages directly on the user’s device.

The company has clarified that during this process, the content of the conversation will not be sent to WhatsApp, any third party or its parent company Meta, which aims to identify potential fraud in a timely manner while maintaining user privacy.