Time has started serving ads to AI agents on markdown pages

Digiday reported on 30 July 2026 that Time has begun serving ads targeted at AI agents rather than human readers, with Ally Bank and the Project Management Institute among the first buyers. The ads run on markdown versions of Time's pages, text-only copies that drop design and images, and are formatted as FAQs carrying a brand's information and messaging.

Time converted all of its webpages into markdown starting last month, around June 2026, reasoning that easier machine access improves its odds of appearing in generative search answers. It works with the AI ad-tech platform Mobian, which converts the pages to markdown and generates the ads. Each ad is labelled as sponsored content at the top, and Digiday notes that no policy currently requires that label. Time and Mobian say this is the first time a publisher has served ads specifically targeting AI agents, a claim that belongs to them rather than an established fact.

How Time and Mobian build and sell an agent ad

Digiday described the mechanics on 30 July 2026. Mobian generates the FAQ ad from a brand brief, exports it as a PDF for human review and client approval, then puts the same FAQ questions to AI search engines and tracks visibility, favourability and accuracy scores over time. Time sells one agent ad per markdown page and charges a premium, on the reasoning that AI-bot impressions on authoritative content are scarce. Mark Howard, Time's chief operating officer, declined to disclose pricing.

Howard told Digiday that Time sees more bot traffic than human traffic on most days, with especially large spikes when a Time100 list launches; Time100 Creators published this month. Digiday cites TollBit data showing Time gets more AI-crawler requests than most of the roughly 7,000 publisher sites in TollBit's network. Howard also said the sales team is pitching agent ads to brands that have already converted their own corporate sites to markdown.

The details Digiday reported on 30 July 2026 sit together like this:

The 15% adoption figure comes from the vendor

Jonah Goodhart, co-founder and chief executive of Mobian, said Mobian's data shows about 15% of brands are powering their own markdown pages. That number is the vendor's own, and Digiday's report carries no methodology, no sample and no definition of what counts as a brand, so read it as a directional claim from a company selling the service rather than a measured market share.

Goodhart's case for the format is about reach: "Maybe it's more important to influence the agent than even the human, because with a human you influence one person. When you influence ChatGPT, you're influencing potentially all of ChatGPT." He also said: "The big idea here is LLMs and AI want trusted information about brands."

What Digiday does not report is whether any of it works. There is no performance data, and the story states it is not known whether LLMs treat these ads differently from editorial content or penalise promotional material. This is a live commercial offering, not a proven channel.

The cloaking risk is the reason not to copy this on your own site

Rob Derow, managing director and partner at BCG X, told Digiday on 30 July 2026 that the biggest risk is the absence of rules on how LLMs treat markdown ads: if LLMs come to view the practice as a form of cloaking, the pages could become less effective or be penalised. That risk is not evenly distributed. A publisher selling a labelled placement on its own markdown pages is in a different position from a brand that puts promotional content into machine-only copies of its own site, where the machine-facing version of a page says something the human-facing version does not.

The ad unit itself is worth one flat observation: what brands are being sold is an FAQ, the same question-and-answer structure answer engines lift from. It fits what the machine is already looking for.

What this means for Thai marketers

Digiday's story names no Asian or Thai market and describes a US publisher's product, so nothing in it is stated to be available in Thailand. The transferable part is what it does to measurement: once publishers can sell placements inside the text that LLMs read, an AI visibility programme has to be able to tell an earned citation from a paid one. A mention retrieved from a page a competitor paid for is a different asset from one earned by being the best source on the question, and a dashboard that counts both as visibility reports a number that can be bought.

For a Thai brand the work still sits on the earned side: making your pages machine-readable, shaping content to answer questions in the form answer engines extract, and keeping entity information consistent enough to be trusted. That is what AI-era SEO already asks for, without the exposure Derow described.

Frequently asked questions

Do ads served to AI agents actually work?

There is no performance data yet. Digiday reported on 30 July 2026 that it is not known whether LLMs treat these ads differently from editorial content or penalise promotional material.

Is this available in Thailand?

The source does not say. Digiday's report covers Time, a US publisher, and names no Asian or Thai market, nor any equivalent offering from a publisher in Thailand.

Should a brand convert its own site to markdown and put ads in it?

Not on the strength of this story. Rob Derow of BCG X told Digiday that the biggest risk is the absence of rules on how LLMs treat markdown ads, and that if LLMs come to see the practice as cloaking, the pages could become less effective or be penalised.

Is Time really the first publisher to serve ads to AI agents?

Time and Mobian claim it is the first time a publisher has done so, and Digiday attributes the claim to them rather than confirming it independently.

If you are building an AI-visibility programme this year, separate earned citations from paid placements in your reporting before the distinction gets harder to see. Get in touch if you want help.