LibraryScenariosWhen AI moderation amplifies invisible bias
Scenario

When AI moderation amplifies invisible bias

Our AI content moderation system keeps flagging posts from marginalized communities as problematic and I can't figure out if it's the training data or if there's bias I can't see.

Aurelius
Online
The coach is replying

More people experience this than they realize.

What we've seen

Your system keeps flagging posts from marginalized groups and you can't tell if it's the data or a bias you can't see.

Ideas that help explain it
Worth thinking about

“Where Are You with AI Across the Enterprise?”

If this sounds familiar, the Library can help you find the bigger picture.