I think the worry is that the questions will be scraped and trained on for future versions.