Effective Use of Clay for Scraping Company Information

I've been using Clay a lot recently to scrape company information of listings (usually just the website which I then fed into other workflows) and it worked pretty well so far. Today I've tried doing it on D&B and noticed that they somehow manage to trick the AI into returning false information (see attached) anyone came across something similar? 🤔 https://www.dropbox.com/scl/fi/m76wowi3zmdtney4jpkxp/Screenshot-2025-04-16-at-4.48.22-PM.png?rlkey=bg81ij5w79gdg98y4azdi3pxz&dl=0