WhatsApp has been taking steps to deal with the fact that its platform is widely used to spread misinformation which has, in some cases, even proven to be fatal. It recently put a limit on text forwarding to prevent rumors from spreading. The Facebook-owned company has revealed that it bans some 2 million accounts every month to fight the spread of misinformation.
The company revealed in a press briefing today that it has built a machine learning system to detect and sort users who showcase specific behavior such as creating multiple accounts and sending messages in bulk just for spreading inaccurate content on the world’s most widely used cross-platform messaging service.
People who use the platform to spread misinformation or create automated fake accounts are banned from WhatsApp at various stages. They may be banned when trying to register, while messaging or when they get reported by other users.
A spokesperson mentioned that over 2 million accounts are banned on WhatsApp monthly. The decisions are based on a machine learning system which uses data on WhatsApp’s previous dealings with unscrupulous accounts and from specific scenarios that engineers followed while taking down accounts.
The company adds that the machine learning system has improved so much that it’s able to ban 20 percent of bad accounts right at the time of registration. The red flags that are considered include the user’s IP address, country of origin of the phone numbers used to sign up, the age of the account, and whether the account started sending messages in bulk soon after it was created, among others.
Filed in . Read more about Whatsapp.