Science and Tech

WhatsApp faces fake news and virality in the elections

Prevention by design

Although WhatsApp cannot influence people’s conversations, Szerman is aware that the platform is usually used to forward chain messages and that is where they have paid attention to prevent the spread of fake news.

The forwarding labels, the spokesperson points out, are intended to mark a message coming from an author that is not the original and in case it has been redistributed more than five times, the message shows a double arrow in the upper corner to indicate an alert that it is highly forwarded content and you should have a critical eye when reading it or sharing it again.

Messages can be forwarded to a maximum of five chats at a time to limit their visibility. “We generate friction and that prevents the viralization of potentially misinformative content,” details Szerman, who maintains that these measures have concrete effects on the amount of viral content that circulates on the platform.

Practices against malicious account use

Despite the limitations, Szerman points out that they have the ability to detect certain behaviors on the part of users to determine that they are trying to create automated accounts in bulk to use them in malicious practices and eliminate them before they can send a message.

In the event that the accounts have been created, some behaviors that give them away are sending many messages one after another. The specialist details that when a real person writes, the legend “writing…” appears under their name, which does not happen with a bot.

“Suppose an account registers and the first thing it does is create groups at a very high scale and at a very high speed, much faster than a human could. Well, that is also abnormal behavior for us and gives us reason to block her,” she comments.

Another way to identify this type of accounts automatically is from the response level, since if it is very low or null it also represents an unusual pattern. Data from the platform highlights that 75% of malicious accounts are detected and blocked without a user sending a report.



Source link