They have admitted to flagging content already - whether the information can be deemed 'harmful' is debatable.

https://www.youtube.com/watch?v=zqEvQKO5_gM