Thanks to Machine Learning, Youtube says it can take millions of YouTube videos that breach its conditions offline before anyone has seen them. In the last quarter of 2017, this happened with 6.7 million videos.

Machines rated these 6.7 million videos as spam or as ‘adult content’ before a human reviewer did this . Of that number, 76 percent was actually removed before a visitor on the video platform could have watched it.

YouTube introduced machine learning to identify offensive videos in June 2017. More than half of videos removed for extreme violence now have fewer than ten views. In early 2017, before the platform started machine learning, eight percent of the violent videos that YouTube removed had less than ten views, the rest had more views. With this example, YouTube wants to show that it is faster in removing videos.

YouTube emphasizes that the use of machine learning means that more people are involved in the review process, because human moderators must rate videos with a flag on the platform’s guidelines. Google aims to employ ten thousand people by the end of 2018 to assess infringing content. Facebook recently announced that it wants to deploy the double number of people for the same task by the end of this year, in addition to security.

In total, YouTube removed eight million videos in the last three months of 2017 due to spam or offensive material. That is a fraction of one percent of the total number of videos added according to the service.

YouTube deleted video Q4 2017