What exactly did Time do that Perplexity objected to?

Time converted its site to stripped-down, text-only markdown pages for AI crawlers and bots — starting in June 2026, a format increasingly common for AI-agent accessibility — and, working with ad-tech partner Mobian, began inserting sponsored, FAQ-formatted brand messaging into those agent-facing pages; the product was already live when Digiday first reported on it on July 30, 2026. [1] Ally Bank and the Project Management Institute were the first advertisers. [1] The sponsored blocks carried a "sponsored content" label, though no industry policy required that disclosure at the time. [1][2]

Does this mean I shouldn't have a markdown or text-only version of my site for AI crawlers?

No — serving a clean, text-only version of your pages to AI crawlers and agents is standard, sensible AEO practice, and it's a different thing from what got Time flagged. The problem wasn't the markdown format; it was injecting promotional, sponsor-paid content into it. A factual llms.txt file or an agent-readable page describing your own products and services is not the same risk category as running third-party paid placements through that channel.

What is a 'trust score' in Perplexity's index, and how would I know if I'd been hit?

Dwyer's warning referenced a "reputational downgrade" and a hit to a publisher's standing in Perplexity's proprietary search index, but Perplexity hasn't published the mechanics, thresholds, or a way for a publisher to check their own status. [2] In practice, the visible symptom would be a drop in how often and how prominently your domain gets cited in Perplexity answers relative to your historical baseline — which is why ongoing citation-frequency tracking across AI engines, not just search rankings, is now a real monitoring requirement.

Is this specific to Perplexity, or should I assume ChatGPT and Google AI Overviews would react the same way?

This is a confirmed Perplexity action only; neither OpenAI nor Google has made a comparable public statement about Time's agent-ads product or markdown-embedded sponsorships as of this writing. Digiday itself frames Perplexity's move as "the latest move in an evolving agentic advertising ecosystem" rather than an isolated event [2] — a reasonable signal that other engines, which depend just as much on user trust in their answers, could take comparable positions as this space matures.

Perplexity Just Blocked a Publisher's Ads From Its AI Agents — What That Means for Your Agent-Facing Content

Q: What did Perplexity actually do, and is it permanent?

Perplexity's chief communications officer Jesse Dwyer confirmed to Digiday on August 11, 2026 that Perplexity is blocking Time's markdown ads from influencing its AI agents and user-facing answers, saying the company works "continuously" to protect users from deceptive practices and that publishers who deploy "deceptive advertising like markdown ads" risk a reputational downgrade. [2] Perplexity did not publish a policy document or detail the exact blocking mechanism, but Digiday separately reported that Perplexity's search and security teams are working on other solutions to address markdown ads more broadly — so this looks like an early application of a developing approach, not necessarily a one-off. [2]

Perplexity confirmed on August 11, 2026 that it is blocking sponsored content Time embedded in the markdown pages it serves to AI crawlers and agents, with chief communications officer Jesse Dwyer telling Digiday that publishers who deploy "deceptive advertising like markdown ads" risk a "reputational downgrade" — including a hit to their trust score in Perplexity's search index.^[2] It's a rare, on-record case of a major answer engine actively penalizing what a publisher puts inside the AI-agent-facing version of its site, rather than just ranking or ignoring it — and it draws a real line for anyone building content specifically for AI crawlers to consume.

Key Facts at a Glance

Detail	Finding	Source
What Time did	Converted its site to markdown pages starting June 2026; began serving those pages with embedded, FAQ-formatted sponsored content — live by the time it was first reported, July 30, 2026	Digiday ^[1]
Ad-tech partner	Mobian, an AI-powered ad-tech platform	Digiday ^[1]
First advertisers	Ally Bank, Project Management Institute	Digiday ^[1]
Perplexity's action	Blocking Time's markdown ads from influencing its AI agents and answers	Digiday, Aug 11, 2026 ^[2]
Perplexity's stated reason	Calls markdown ads like Time's "deceptive advertising"; says it works "continuously" to protect users from deceptive practices	Jesse Dwyer, Perplexity CCO ^[2]
Stated consequence for publishers	"Reputational downgrade" and a hit to trust score in Perplexity's index	Jesse Dwyer, Perplexity CCO ^[2]
What's still undisclosed	Exact blocking mechanism, how "deceptive" is defined, whether other publishers are affected	Digiday ^[2]

What Did Time Actually Build?

Time converted its site into stripped-down, text-only markdown pages starting in June 2026 — a format increasingly common for improving how AI systems parse and cite a site's content. Working with ad-tech partner Mobian, Time then began inserting sponsored content into those same agent-facing pages, formatted as "FAQs with a brand's information and messaging, and labeled as sponsored content."^[1] Ally Bank and the Project Management Institute were the first buyers.^[1] No industry standard required that sponsored-content label at the time.^[1][2]

Time's chief operating officer Mark Howard framed it as a natural extension of existing sponsorship formats: "This is an obvious extension of list sponsorship or franchise sponsorship, because we do know that there is a spike of the agents coming."^[1] Mobian co-founder and CEO Jonah Goodhart made the underlying commercial logic explicit: "When you influence ChatGPT, you're influencing potentially all of ChatGPT. If ChatGPT changes what it says about [a brand], it's massive and it's more than any one campaign could ever do."^[1] Howard also acknowledged the experiment's novelty and uncertainty: "We don't know yet because this is brand new, and we believe that we are paving the first path forward here."^[1]

That candor turned out to matter. Less than two weeks after Digiday first reported the product was live, on July 30, 2026, one of the AI engines it was built to influence pushed back.

What Did Perplexity Do, and Why?

Perplexity confirmed to Digiday that it is blocking all markdown advertising from Time.com from influencing its agents and the answers it surfaces to users.^[2] Dwyer's statement drew a direct line between the practice and user trust: Perplexity works "continuously" to protect users from deceptive practices, and publishers who deploy "deceptive advertising like markdown ads" risk "a reputational downgrade" and damage to their standing in Perplexity's proprietary trust scoring.^[2]

Two things are worth being precise about, because the coverage of this story has been thinner than the headline suggests. First, Perplexity's confirmation came through an emailed statement to a reporter, not a published policy document or blog post — there's no public rulebook yet describing exactly what triggers a downgrade, only a named executive's on-record account of what the company did and why.^[2] Second, Perplexity did not answer follow-up questions about how it technically identifies or blocks the sponsored blocks. Digiday did separately report that Perplexity's search and security teams are working on other solutions to protect users from markdown ads "more broadly" — evidence this may be an early application of a developing approach rather than a one-off reaction to Time specifically.^[2] Time did not respond to Digiday's request for comment on Perplexity's action.^[2]

That gap between "a real, on-record enforcement action happened" and "a fully documented, generalizable rule exists" is exactly the kind of thing worth tracking rather than over-interpreting. What's solid: a named Perplexity executive, on record, confirming the company is actively filtering what it treats as manipulative content inside agent-facing pages, and attaching a stated consequence — a trust-score hit — to publishers who try it.

Why This Matters If You're Building Content for AI Agents

The instinct to read this as "don't make separate content for AI crawlers" would be the wrong takeaway. Serving a clean, structured, text-first version of your site to AI agents — the same principle behind an llms.txt file or an agent-readable summary of your offerings — remains sound practice; it's how you make yourself easy and unambiguous to cite in the first place.^[3] What got flagged wasn't the existence of an agent-facing channel. It was using that channel to run paid, promotional insertions dressed as neutral facts, aimed at a system's judgment rather than at informing it.

That distinction should shape how you think about any agent-facing content you maintain:

Facts about your own business are not the risk category. A markdown page, an llms.txt entry, or an agent-readable FAQ describing your own products, pricing, or service scope in your own voice is standard AEO infrastructure, not the practice Perplexity flagged.
Third-party paid placements inside agent-facing pages are the risk category. If you're a publisher weighing whether to sell "agent ad" inventory the way Time did, understand that at least one major answer engine has now demonstrated it will actively police and penalize that model — not just decline to surface it.
Labeling alone didn't protect Time. The sponsored blocks were disclosed at the top of each page, and Perplexity still called markdown ads like Time's "deceptive advertising."^[2] A visible label is not, on its own, evidence that content wasn't judged manipulative by the system reading it.
This is a live precedent, not settled law. Perplexity hasn't published a policy, and no other major engine has taken or announced a comparable public position on Time's product as of this writing. Treat the trust-score warning as a real signal about enforcement appetite, not a finalized rulebook you can audit against line by line.

What Should You Check on Your Own Site This Week?

Audit any agent-facing or markdown version of your site for embedded ad tech, sponsored copy, or third-party promotional insertions you may have added — or that an ad-tech vendor may have added on your behalf — separate from what human visitors see.
Keep your llms.txt and any agent-readable pages first-party and factual. Describe your own offerings plainly; don't use that surface to carry paid placements for other brands, even clearly labeled ones. See our full breakdown of what llms.txt actually is and who reads it if you haven't set one up yet.
Track citation frequency, not just rankings. A trust-score downgrade in an answer engine's index wouldn't show up in classic SEO tools — the visible symptom is fewer or less prominent citations in that engine's answers over time. Our Multi-Engine Visibility platform tracks exactly this across ChatGPT, Perplexity, Gemini, and Google AI Overviews/AI Mode, the same engines covered in how ChatGPT decides what to cite.
If you're a publisher considering agent-ad inventory, wait for a published policy before committing. Time's team called this "brand new" territory even as they launched it;^[1] Perplexity's response confirms the downside risk is real and immediate, while the upside case is still unproven.
Understand this sits inside the broader shift already covered in GEO vs. SEO: optimizing for an AI answer engine now includes not just what you publish, but how that content behaves once a system decides whether to trust it.

The bigger pattern here isn't about one publisher's ad experiment. It's that answer engines are starting to actively adjudicate the content inside the pages they retrieve — not just rank or ignore it — and at least one of them has now shown it's willing to name a specific practice as deceptive and attach a consequence to it in public. If your AEO strategy involves any agent-facing content beyond your own factual pages, that's a new variable worth planning around now, before a specific practice you're running gets named the same way.

See where you stand in AI search. Run a free AEO audit of your site — free with a 3-month plan minimum.

References

[1] Digiday, "Time has started serving ads to AI agents," July 30, 2026. https://digiday.com/media/time-has-started-serving-ads-to-ai-agents/ — confirms Time began converting pages to markdown in June 2026, the Mobian partnership and ad format ("FAQs with a brand's information and messaging, labeled as sponsored content"), that the product was live by this article's July 30, 2026 publish date, named advertisers (Ally Bank, Project Management Institute), and on-record quotes from Time COO Mark Howard and Mobian co-founder/CEO Jonah Goodhart.

[2] Digiday, "Perplexity blocks Time's ads served to AI agents, calling them 'deceptive,'" August 11, 2026. https://digiday.com/media/perplexity-blocks-times-ads-served-to-ai-agents-calling-them-deceptive/ — confirms Perplexity's blocking action, the on-record statement from chief communications officer Jesse Dwyer ("continuously" protecting users; "deceptive advertising like markdown ads"; "reputational downgrade"/trust-score language), that Perplexity's search and security teams are working on broader solutions, and that Time did not respond to a request for comment.

[3] RankGarage, "llms.txt: What It Is, Who Respects It, and How to Write One," August 8, 2026. /blog/llms-txt-guide — internal reference for standard, non-promotional agent-facing content practice.