Plot twist: those instructions are written with white text on white background and serve to eliminate AI-generated resumés.
I mean, LLMs are also able to respond to HTML and CSS that indicates hiding prompts, if you wanna parse the DOM when doing this, which I would. And manually double or triple check every outgoing submission…