Time started serving advertisements to AI agents this month, and the ads aren’t banner images or video, they’re sponsored FAQ-style text blocks written specifically for a machine to read. Ally Bank and the Project Management Institute are among the first brands to buy one. The logic is straightforward once you see the underlying number: Time says it now sees more bot traffic than human traffic on most days, which lines up with broader web data showing bot traffic has overtaken human traffic across the internet generally.
The mechanism behind this: Time converted its site into stripped-down markdown versions of its pages last month, text-only, no design, no images, built specifically to be easy for AI systems to crawl and process. Working with an ad-tech platform called Mobian, Time then generates agent-targeted ads from a brand brief and drops them into those markdown pages alongside the real content.
The obvious question nobody in this story has answered yet: do AI models actually recognize these as ads, or do they just read them as more content to summarize and repeat as fact? No major AI company has said how its models handle ads embedded in markdown pages, and that ambiguity is exactly the risk. If AI companies eventually treat this the way Google treated cloaking, serving different content to crawlers than to humans, these pages could get penalized rather than rewarded — the same fate that hit the SEO tactics this setup structurally resembles.
There’s a bigger shift under this one story: publishers spent two decades optimizing for Google’s crawler. Now some are starting to optimize for a different set of crawlers entirely, the ones that read a page once, summarize it, and never send a human visitor back to the original site at all. Whether “selling ads to bots” turns into a real revenue line or a cautionary tale probably depends on an answer only the AI labs can give.
Time isn’t the only publisher rethinking what its content owes an AI-driven audience versus a human one. Meta’s new mandatory AI ad disclosure rules are the mirror image of this story: one company is trying to make sure humans know when they’re looking at AI-made content, while Time is trying to make sure AI systems notice a human-facing ad at all. Both are early, uncertain attempts to write rules for a web that didn’t have this problem three years ago. Digiday’s original reporting goes deeper into how Time built the markdown pipeline itself.