67
New research has found that AI-powered content moderation systems from Google, OpenAI, Anthropic, and DeepSeek don’t always come to the same conclusions about bad language on the internet.