Time Magazine has launched an advertising format visible only to AI crawlers — bots from ChatGPT, Claude, and Perplexity receive sponsored content on a separate version of the site, inaccessible to live readers. German developer Vincent Schmalbach discovered that Anthropic and OpenAI crawlers load pages with Ally Bank integrations — FAQs about banking products and marketing talking points — while regular visitors and Google's search bot see editorial content without ads. The technology is called "agent ads" and was created by adtech company Mobian: the goal is to influence AI assistant responses at the moment they build their knowledge base.
How advertising for AI crawlers works
Time has been converting web pages into machine-readable format since June 2026. When ClaudeBot or OAI-SearchBot requests an article, the server delivers HTML with an embedded sponsored block — structured data about the product, benefits, and answers to common questions. A regular visitor receives the same article, but without the commercial layer. Verification using crawler simulators, conducted by The Register, confirmed: content for bots and humans differs at the source code level.
The first advertisers were Ally Bank and Project Management Institute. Mobian CEO Jonah Goodhart explains the logic: "When you influence ChatGPT, you influence potentially all ChatGPT users. If the model changes what it says about your brand, that's bigger than any campaign for live people — with a person you influence one individual, with AI you influence millions of queries."
How this differs from GEO
Generative search optimization (GEO) increases the probability of a site being cited in AI responses through content structure, Schema.org markup, and source authority. Agent ads go further: the publisher substitutes content for the crawler, adding a commercial message that the model indexes as part of editorial material. A user asks a question — the AI synthesizes an answer based on indexed data, including the sponsored block, but doesn't disclose that part of the information is paid.
In search advertising, ads are labeled "Ad," on social media integrations are marked "#ad" or "Sponsored." Agent ads currently have no such labeling: neither the crawler nor the end user receives notification of the commercial nature of the data. This creates a risk to audience trust and potential regulatory claims.
Influencer marketing in the age of AI intermediaries
Until now, influencer marketing has been built on direct brand-to-audience contact through bloggers: reach was measured by views, engagement by reactions and comments, conversion by link clicks. Agent ads shift the point of contact to the AI assistant level: the brand influences not the human directly, but the source from which the human gets a recommendation. If a user asks ChatGPT "which bank should I choose for savings," the model may cite Ally Bank data indexed via Time's agent ads — and the user perceives this as neutral advice, not advertising.
Brands gain an influence channel that works like word-of-mouth at the scale of millions of queries — but without explicit attribution and the ability to trace the source of the recommendation.
For the Russian market, this means a need to revise media plans. If Western publishers scale agent ads, and Yandex or local AI services implement similar formats, brands will face dual optimization: one strategy for live readers and bloggers, another for AI crawlers. Meanwhile, Russian legislation requires explicit ad labeling: if an AI assistant cites sponsored material without indicating its commercial nature, the question of liability arises — does it fall on the publisher, the model developer, or the advertiser.
Risks and format limitations
Agent ads are opaque to the end user. In Google search results, ads are highlighted with color and a label; in a blogger's feed, integrations are marked by law. Here the user doesn't know that part of the AI response was generated based on paid content. This increases message effectiveness — no banner blindness or advertising skepticism — but reduces trust if the mechanics become public.
Measuring results is difficult. In influencer marketing, brands get reach stats, engagement rates, and UTM click-through data. Agent ads don't provide direct attribution: it's impossible to know how many times ChatGPT cited Ally Bank's sponsored block or how many users made decisions based on that information. Mobian and Time have not yet disclosed campaign effectiveness metrics.
The competitive landscape will change quickly. If the format proves profitable, other publishers will launch similar products, crawlers will start receiving sponsored content from dozens of sources, and AI models will face conflicting data — one publisher promotes brand A, another brand B. How the model decides what to cite depends on ranking algorithms that AI developers keep proprietary.
Practical checklist for marketers
If your brand is considering placement in agent ads or preparing for such formats to appear on the Russian market, use the following sequence of actions:
- Audit your current presence in AI responses: ask ChatGPT, Claude, Perplexity about your product or category — record which sources the model cites and how it describes the brand.
- Competitor analysis: check whether competitors are mentioned in AI responses more frequently and in what context — this will show who is already optimizing content for crawlers.
- Prepare structured data: create FAQs, product descriptions, and key talking points in machine-readable format — even without purchasing agent ads this will improve indexing for GEO.
- Legal review: clarify with legal counsel the requirements for ad labeling in the context of AI assistants — Russia's Advertising Law applies, and implicit promotion may result in fines.
- Test attribution: if launching a pilot with agent ads on foreign platforms, use unique promo codes or UTM parameters in AI responses (if the format allows) — this will help you evaluate conversion.
- Monitor reputation: track brand mentions in AI responses regularly — if the model starts citing outdated or negative data, you'll need to quickly update your content.
Frequently asked questions
Can agent ads be used in Russia
As of August 2026, the agent ads format is available only through Time Magazine and Mobian, which work with Western AI models. Russian publishers and adtech platforms have not yet launched similar products. If the format appears on the local market, it will need to comply with the Advertising Law — explicit labeling of commercial content is mandatory, otherwise fines are possible for the advertiser and platform.
How to measure the effectiveness of advertising for AI crawlers
Direct attribution is currently absent: AI model developers don't disclose how frequently the model cites a specific source. Indirect metrics — growth in branded search queries, increased website traffic from new sources, dynamics of brand mentions in AI responses (tracked manually through regular queries to the model). For accurate assessment, APIs from OpenAI, Anthropic, or partnerships with adtech platforms providing analytics will be needed.
Will agent ads replace influencer marketing
No, the formats solve different problems. Influencer marketing works on engagement, emotional connection, and trust through the blogger's personality — the audience sees the product in context, asks questions, shares opinions. Agent ads influence the information layer: when a user seeks a fact or comparison, the AI gives an answer based on indexed data. Brands will use both channels in parallel: bloggers for reach and loyalty, agent ads for influencing the decision-making moment through AI assistants.
In brief
- Time Magazine sells advertising visible only to AI crawlers from ChatGPT, Claude, and Perplexity — sponsored content is embedded in pages inaccessible to live readers.
- The agent ads format differs from GEO: instead of optimizing existing content, the publisher substitutes content for bots, adding a commercial message.
- First advertisers are Ally Bank and Project Management Institute; the technology was created by adtech company Mobian.
- The main risk is the lack of labeling: users don't know the AI response contains advertising, which contradicts transparency principles and may violate advertising law.
- For the Russian market, the format will require adaptation to comply with the Advertising Law and creation of attribution tools — currently direct effectiveness measurement is impossible.
- Marketers should start auditing their brand's presence in AI responses, prepare structured data, and track how competitors optimize content for crawlers.
Planning a campaign with creators or adapting your strategy for emerging channels? ETC helps you build media buying and reach analytics that account for how audiences actually consume content today.
