“One thing we see offenders try to do is connect with children on one platform and move somewhere they can have a private communication or private video stream,” said Lindsey Olsen, a former executive director for the NCMEC’s exploited children division.

Algorithms can scan websites and apps for abusive material by comparing images and videos to content that’s already been identified as inappropriate. Livestreams showing abuse are brand new, so it’s hard for companies to quickly identify them as inappropriate and remove them, said Rebecca Steinbach, senior producer at NCMEC.

I feel like AI should be able to detect them now, without having to just use the database they created 2009.