Ads have always found a way to show up wherever people are, and now they are taking a surprising step: targeting robots directly. Time magazine just announced something that sounds like it came straight out of a science fiction script, but it is very real — the publication has started running ads designed specifically for AI agents, not humans. Ally Bank and the Project Management Institute were the first to jump on board this initiative, developed in partnership with the ad tech platform Mobian, and the move is already turning heads across the entire digital media market.
What seems odd at first glance makes a lot of sense once you understand the context. Last month, Time began converting all of its pages from HTML to markdown, a plain text format without images or heavy visual elements that AI systems and agents can read with far more ease and accuracy. The original plan was to improve visibility in generative search results — those answers that pop up directly inside tools like ChatGPT or Gemini, without the user ever needing to click a link. What nobody expected was that bot traffic would grow so much it would surpass human visitors on most days. And so, like any media outlet would, Time spotted a real monetization opportunity in this new reality. 🤖
Why Markdown became the centerpiece of this strategy
Markdown is a text markup format originally created to make writing for the web easier, but it has taken on a whole new purpose with the rise of language models. Unlike traditional HTML, which carries a massive amount of tags, scripts, images, and visual elements, markdown delivers content in a clean, structured, and extremely machine-readable way. When an LLM needs to process a page, it does not care about the visuals — it wants the text, the information hierarchy, and the context. Markdown delivers exactly that, with no noise, no unnecessary bulk, and a structure that models can absorb far more efficiently than any other format available today.
Time’s decision to migrate its content to this format was not just technical — it was strategic. The publication realized that a significant portion of its traffic was no longer coming from human readers typing the site address into a browser, but from automated systems, crawlers, and AI agents sweeping the internet for information to feed answers. Worth noting: there is still no consensus on how effectively markdown pages can actually serve as a barrier against declining search traffic. Placing ads on those pages is precisely the way Time found to squeeze some revenue out of growing bot traffic while it waits to figure out the answer.
To make all of this work, Time partnered with Mobian, an AI-powered ad tech platform that converts pages to markdown and generates what are being called agent ads. These ad units are formatted as questions and answers, FAQ-style, containing the brand’s information and messaging, and they are clearly labeled as sponsored content. According to Time and Mobian, this is the first time a publication has run ads with the specific goal of reaching AI agents. And from the looks of it, plenty of others are already lining up.
Receive the best innovation content in your email.
All the news, tips, trends, and resources you're looking for, delivered to your inbox.
By subscribing to the newsletter, you agree to receive communications from Método Viral. We are committed to always protecting and respecting your privacy.
How monetization for AI agents actually works
The logic behind this monetization strategy is more powerful than it might seem at first. When an LLM like ChatGPT, Claude, or Perplexity absorbs ad information while reading a page, it can start including that information in the responses it generates for users. This means a single ad unit, well-positioned and properly formatted in markdown, could theoretically influence a lot of people at the same time. The ad does not appear as a banner, it does not flash, it does not interrupt — it simply becomes part of the information the model uses to build its answers.
Jonah Goodhart, co-founder and CEO of Mobian, sums up this shift in mindset well. According to him, it may be more important to influence the agent than the human, because with one person you impact just one individual, while by influencing ChatGPT you potentially impact all of ChatGPT. If the tool changes what it says about a brand, the effect is massive — bigger than any standalone campaign could ever achieve. In his view, the key insight is that LLMs and AI in general want reliable brand information, and Mobian delivers that in a direct and easy-to-consume format.
Mobian’s own data shows that roughly 15% of brands already maintain their own markdown pages, and the trend is for that number to grow as more companies bet on a two-layer internet — one for humans and one for machines. Mark Howard, Time’s chief operating officer, shared that the sales team is pitching these ads specifically to brands that have already converted their corporate sites to markdown, because those are the ones already thinking about how to reach AI bots. Agent ads would be yet another way to get in front of language models.
Ally Bank and the Project Management Institute, a nonprofit professional organization, gain a new lever to control how they appear in generative search. LLMs pick up the ad information and index it, which can shape AI-generated answers, which in turn influence the humans using those response tools. That is why, internally, these ads are seen as part of a broader effort by Time to sell a GEO (generative engine optimization) product to brands, built on what the publication has learned about where AI search engines look for information and when Time shows up in those summaries.
How the ads are created and measured
One interesting point is that the production process is not all that complicated. Mobian can generate a brand ad in FAQ format from a briefing. That piece is converted into a PDF so humans can review it, then shared and approved by the client — a workflow that looks a lot like a traditional branded content ad. After that, Mobian submits those same questions to AI search engines to measure the ad’s performance, tracking metrics over time such as visibility, favorability, and information accuracy.
These ads can be contextually targeted, tied to Time’s editorial lists and franchises, or set by date range. And they are not cheap, either. Time charges a premium for these units. The reasoning is that advertisers are paying for the chance to shape the information LLMs retrieve, since AI bot impressions on authoritative content are scarce and valuable — and it is precisely that scarcity that opens the window to influence responses. The publication sells only one agent ad per markdown page and measures bot traffic on that page, although Howard preferred not to disclose the pricing.
Bot traffic has already surpassed human traffic
Howard did not want to share the exact numbers behind this traffic increase, but he cited data from TollBit showing that Time receives more AI crawler requests for its content than most of the nearly 7,000 publication sites on TollBit’s network. That spike in bot traffic becomes even more obvious when Time launches one of its Time100 franchises, like the Time100 Creators list published this month. There is so much automated access that, on most days, the publication logs more bot traffic than human traffic — something that mirrors Cloudflare data indicating that more than half of all web traffic today already comes from bots.
In Howard’s view, this is a growing traffic source and, therefore, a growing source of ad inventory with enormous potential value — as long as Time can keep developing the product to the point where it has a consistent offering for brands. He sees agent ads as a natural extension of list and franchise sponsorships, since there is a predictable spike in agents accessing that content. The ability to align contextually, combined with domain authority, volume, and the ability to schedule and target campaigns, is all coming together in a way that, according to him, should become a very important component of Time’s offering to brand partners.
The risks and unknowns of this new frontier
Of course, not everything is certain in this uncharted territory. It is still unclear whether LLMs treat ads differently from editorial content, or whether they penalize promotional material when deciding what information to retrieve — after all, this has practically never been done before. But there is precedent for tightening rules around where and how ads can appear: Google began enforcing stricter standards and anti-spam policies for ads on publisher sites throughout the 2010s.
Rob Derow, managing director and partner at BCG X, the technology and design unit of the Boston Consulting Group, points out that the biggest risk with markdown ads lies precisely in the absence of rules governing how LLMs handle them — and in how that could change down the road. If models eventually view this practice as a form of cloaking, that SEO tactic where a site serves different content to crawlers and to humans, publications could see these pages lose effectiveness or even get penalized for including sponsored content in markdown files. Still, Derow notes this can be easily monitored with AI visibility tools.
With that in mind, Time’s ads include a sponsored content label right at the top that names the client brand, even though no policy currently requires that kind of disclosure. Howard and Goodhart emphasize that ads are already being served to AI agents when they retrieve information from publisher pages, so Time chose to get ahead of the curve. As Howard himself put it, nobody knows for sure what lies ahead, because all of this is brand new — and they believe they are paving the first path in this direction, applying lessons learned from how they have handled this kind of advertising in the past.
What this means for the future of digital advertising
This move by Time is not an isolated experiment — it is a clear signal of where the market is heading. As more people start using LLMs as their first stop for finding information, whether to research products, compare services, understand concepts, or make decisions, traditional organic traffic to sites and publications will keep shifting. Fewer humans arriving through conventional search, more AI agents sweeping content to power answers. Publications that do not adapt to this new reality risk watching their relevance shrink — not because the content got worse, but because it simply stopped being accessible to the systems that now distribute information at scale.
For brands and advertisers, the message is clear: the concept of digital monetization is being rewritten right now, and those who figure out first how to create advertising content that is genuinely useful and readable for AI systems will have a serious head start. It is no longer just about showing up for humans at the right moment — it is about being present in the knowledge layer that models use to build their answers. Markdown is just the beginning of this story, and what Time is doing today is probably the first page of a chapter that will redefine how we think about reach, audience, and influence in the age of language models. 🚀
