Model welfare is an emerging field of research that seeks to determine whether AI is conscious and, if so, how humanity should respond.