YouTube has been working on ways to manage offensive and extremist content that do and do not violate its policies, and some steps it has taken include AI-assisted video detection and removal as well as input from more experts. Today, in a blog post, the company provided more detail about its ongoing efforts.
First, its machine learning video detection has been hard at work and during the past month over 75 percent of videos taken down because of violent, extremist content were done so without the help of humans. This system has helped YouTube remove twice as many of these sorts of videos. The company has also started working with a number of non-governmental organizations including the Anti-Defamation League, the No Hate Speech Movement and the Institute for Strategic Dialogue. "These organizations bring expert knowledge of complex issues like hate speech, radicalization, and terrorism that will help us better identify content that is being used to radicalize and recruit extremists," said YouTube in the blog post.
For videos that contain "controversial religious or supremacist content" but don't violate any of YouTube's policies, they'll now be placed in a "limited state." YouTube said, "The videos will remain on YouTube behind an interstitial, won't be recommended, won't be monetized, and won't have key features including comments, suggested videos, and likes." It says that the limited state will start being applied to desktop versions in the coming weeks and will hit mobile versions shortly thereafter.
YouTube said that these changes are just the beginning and it will be sharing more about its work in the months ahead. "Altogether, we have taken significant steps over the last month in our fight against online terrorism. But this is not the end. We know there is always more work to be done," it said.