How does Google/FB identify offensive content?

[thcricketfan]

After the NZ shooting, Google/FB struggled to identify offensive content from YT. I am wondering what algo/system design has put in place to identify offensive content. It is not an easy problem to solve, especially if it is a live-stream Additionally, how do they identify duplicate or content that is similar to something that has been flagged previously. Thoughts?

[LinuxBender]

Offensive content, or content that violates the acceptable use policy? Everything is offensive to someone.

They have a reporting function to flag violations, that apparently nobody used until the live stream completed. [1]

[1] - https://newsroom.fb.com/news/2019/03/update-on-new-zealand/