One of the most comprehensive studies to date on the effectiveness of online hate-speech filters underscores a shortcoming many social-media users know firsthand: They don’t always catch hateful speech.
The study highlights the difficulties of training artificial-intelligence algorithms in rooting out slurs from social media and other sites without blocking legitimate posts that contain similar language—whether by news organizations, groups opposed to hate speech, or racial groups that have appropriated those terms.
“Detecting…