The Center for Countering Digital Hate tested the safety guardrails of popular chatbots and found them lacking.
