Picture this: you ask an AI assistant to find you the best noise-canceling headphones under $200. It reads a dozen review sites, compares specs, and gives you three solid picks. You never visit a single website. You never see a single ad.
So who paid for all those reviews to get written?
That's the question publishers are scrambling to answer right now, and the answer some of them have landed on is strange: invisible ads. Ads built not for your eyes, but for the AI reading the page on your behalf.
The Problem AI Agents Created
For twenty years, the internet ran on a simple deal. Publishers wrote stuff, readers showed up, ads ran next to the content, and everybody got paid enough to keep going.
AI agents just broke that deal.
When ChatGPT, Perplexity, or some shopping assistant crawls a page to answer your question, it's reading the content without you ever landing on the site. No page view. No banner ad seen. No affiliate click. The publisher did the work, the AI got the value, and the human reader got the answer — all without the website ever collecting a dime.
I run into this constantly when I'm researching a trip. I'll ask an AI tool to compare visa requirements for three countries or find me the cheapest route between two cities, and it'll pull together an answer from a half-dozen sites in seconds. Useful for me. Rough deal for whoever spent hours writing the original guide.
The Invisible Ad Fix
So some publishers are trying something clever, maybe a little sneaky: embedding ad content in a page's code that's invisible to a human scrolling through, but fully visible to the AI agent crawling it.
The idea is that if an AI is going to "read" your page and summarize it for a user, it might as well also read the sponsor message, the product recommendation, or the affiliate link baked in alongside the content. The human never sees it. The AI does — and maybe passes some version of it along in its answer.
It's a workaround for a system that wasn't built with AI agents in mind. And it raises a question I can't quite shake: is this any different from the old black-hat SEO trick of cloaking, where you showed Google's crawler one page and a human visitor another? That used to get sites banned. Now it's being treated as a survival strategy.
Why This Matters to You
Here's the part that should make you pause. If AI agents are reading things you'll never see, how do you know what's actually shaping the answer you get?
I've spent the last eight years traveling full-time, and most of my trip planning now involves some AI tool at the start of the process. I like the speed. I like that I can ask a dumb question at midnight and get a decent answer. But I also like knowing where information comes from. That's half of what curiosity even means — not just getting an answer, but understanding how you got it.
If invisible ads start shaping what AI agents recommend, that transparency gets murkier. You might think you're getting an objective comparison of travel insurance policies, when really the AI absorbed a sponsored nudge you'll never lay eyes on.
For publishers, I get the logic. If your writing is getting read by machines instead of humans, you need a way to get paid by machines instead of humans. Fair enough. But it only works long-term if it stays honest. The moment readers find out they're being steered by content they can't see, trust disappears — and trust is the one thing no business model survives without.
What Happens Next
This is still early. Nobody's figured out the "right" way to monetize an internet where AI does the reading and humans just want the conclusion. Publishers are experimenting. Some of it will work. Some of it will blow up in their faces once people notice.
My advice? Stay curious about what's happening behind the page, not just on it. Ask your AI tools where their answers come from. Click through once in a while. The internet's always finding new ways to reinvent itself — the trick is making sure you're still the one steering, not the algorithm quietly doing it for you.
Source: The Information
Read our previous article: California Just Told AI: Stop Pretending to Be Human
