Crawlers already know how to stop crawling recursive or otherwise excessive/suspicious content. They've dealt with this problem long before LLM-related crawling.