The company introduced safeguards after readers flagged “bigoted” language in an artificial intelligence feature that crafts summaries.

The company introduced safeguards after readers flagged “bigoted” language in an artificial intelligence feature that crafts summaries.