Facebook CEO Mark Zuckerberg defended Facebook's content moderation policies in an "Axios on HBO" interview, noting that the company proactively removed roughly 90% of hate speech content from April to June this year.
What he's saying: It's more difficult to train AI to identify videos of racist attacks, and to understand the intent behind those videos, than to delete posts that promote terrorism or nudity, Zuckerberg said.
- "Are they posting that video with an intent to encourage people to copy that act and carry out their own racist attacks? Or are they condemning it? Because if they're condemning it, then maybe that should be allowed. But if they're trying to encourage people to copy it, then that's obviously an issue."