Outlets like The Guardian and The New York Times are scrutinizing digital archives as potential backdoors for AI crawlers.