WhatsApp Turns to Artificial Intelligence to Identify Scams
WhatsApp is testing a new feature that can warn users about potential scams before continuing a suspicious conversation.
This feature, called Scam Alert, is optional and uses a machine learning model on the device to analyze messages received from people who are not in the user's contact list. The model is trained on common patterns in scam conversations and analyzes the structure of the dialogue and linguistic cues to identify the likelihood of a scam.
If a suspicious message is detected, the alert is displayed only to the user, and the other party does not see it. The user can then choose to continue the conversation, block the person, or report them.
Meta has emphasized that these checks are performed on the device itself and that message content is not sent to WhatsApp servers for scam detection. Additionally, no reports are automatically recorded, and sending message content or even information related to scam identification only occurs if the user chooses to do so.
Users can also mark a conversation mistakenly identified as suspicious to remove the alert, and this chat will no longer be flagged by Scam Alert. In this case, there is also the option to send the last five received messages to WhatsApp to help improve the accuracy of this feature.
The Scam Alert feature is currently being tested on a limited scale in beta, and Meta has not yet announced the public release date for all WhatsApp users.
Source: Gulf Times
نظرات
حداقل ۳ کاراکتر؛ پس از تأیید مدیر نمایش داده میشود.
هنوز نظری ثبت نشده است
