YouTube and Facebook are working on auto-blocking extremist content

YouTube and Facebook are said to be working on systems to automatically identify and block extremist content, such as videos from the Islamic State, in the same way as copyrighted material is now, for example.

The companies are not publicizing the news officially, possibly because their determination of what will and will not fall under ‘extremist content’ in the eyes of the public or other parties could be at odds with freedom of expression. The information comes out through anonymous Reuters sources.

At the moment, YouTube and Facebook still rely on the ‘human hand’ when it comes to fighting extremist content. Users report to YouTube and Facebook, after which employees of those companies decide whether the content violates the terms of use or not. In the future, these systems should be automated as much as possible: videos labeled as “not acceptable” are assigned a hash, similar to what is now done with obviously illegal material, such as child pornography or illegal copies of copyrighted work. Based on that hash, reposts of the extremist content, even if they have been modified, can be automatically and quickly identified and removed.

The moves come after US President Obama and European leaders urged companies to cooperate in combating extremist content from the Islamic State, for example. For example, the US has already proposed a system to nip content in the bud as much as possible, but Alphabet and Facebook have never been served by government control. Instead, the companies see more in a jointly set up organization that takes on the task of formulating policy and strategy. Whether that will actually happen is unknown at this time. According to the sources, YouTube and Facebook would in any case start hashing what they believe to be extremist content.