WhatsApp blocks two million accounts a month
In this fact, it is not interesting how much you ban, but how!
In the case of social platforms, one of the most pressing problems these days is to filter out fake accounts that are used to spread spam or other illegal activity. WhatsApp executives spoke about the problem and of course the solution at a press conference in New Delhi.
According to them, the company has developed a machine learning system that detects and disables accounts that are “behaving inappropriately”. Inappropriate behavior includes sending bulk messages and creating multiple accounts to distribute “questionable” content. The machine learning system has reached the point where it can segregate 20 percent of non-compliant accounts in the registration phase.
Matt Jones from WhatsApp said the system looks at various factors before banning the connection. They use a list that includes, among other things, IP addresses and phone numbers, ones that have already been used to create similar accounts. That, of course, would not be enough, with artificial intelligence being able to detect 75 percent of disabled accounts per month in XNUMX percent of cases without human intervention.