The conversation around Anthropic's issues with the U.S. government underscores the need for a balanced approach to AI regulation that protects both innovation and public safety.