Anthropic proposes AI policy recommendations to the White House, focusing on safety, security, and power for AI development.
