18 Countries, Including US and UK, Release Guidelines for Enhancing AI Security and Development