Anthropic says it blocked users who tried to use its AI models for potentially harmful biological research, including work involving viruses and toxins.
