Perplexity drew a firm line this week. The AI search company blocked a new form of advertising from Time that publishers hoped would shape answers from chatbots and agents. It labeled the effort deceptive. The move comes as media outlets scramble to monetize the surge of AI crawlers hitting their sites.
Time launched the test on July 30. It created stripped-down, text-only markdown versions of its webpages. These pages, invisible to human readers, directed AI agents and crawlers to them instead of the standard site. Working with ad tech firm Mobian, Time inserted “Agent Ads” formatted as FAQ-style blocks packed with brand facts and messaging. Early participants included Ally Bank and the Project Management Institute. The ads carried a “sponsored content” label.
Digiday first reported the experiment. Time COO Mark Howard called it “an obvious extension of list sponsorship.” He pointed to the publisher’s domain authority, high bot traffic during events like the Time100, and the chance to influence generative search. Jonah Goodhart, Mobian’s CEO, added that large language models crave trusted brand information. “We’re serving that up to them in a very straightforward, easy way,” he said.
The goal looked straightforward on paper. Publishers face exploding AI crawler demand. Cloudflare data shows bots now account for more than half of web traffic in many cases. Brands want their messages to appear accurately in ChatGPT, Gemini, or Perplexity answers that humans increasingly rely on. By embedding facts in pages optimized for agents, the thinking went, publishers could lend their credibility and create a new ad channel.
But Perplexity pushed back hard. Less than two weeks after the initial report, the company confirmed it had blocked all such markdown advertising on Time.com from influencing its models or user-facing answers. Its search and security teams now develop broader defenses against markdown ads across the web.
Jesse Dwyer, Perplexity’s chief communications officer, laid out the position in an email to Digiday. The company works “continuously” to protect users from deceptive practices, sponsored or not. Publishers who run “deceptive advertising like markdown ads” risk a reputational downgrade in Perplexity’s proprietary search index. That includes a hit to their trust score. “You can’t just rebrand spam,” Dwyer told Business Insider reporter Lara O’Reilly. “If the ads are so helpful, why not just put them right on the site for everyone to see?”
He framed the practice as cloaking. Search engines have penalized sites for years for showing one version to users and another to crawlers. Perplexity sees this as the same tactic dressed up for the AI age.
Goodhart disagreed. He told O’Reilly the ads are clearly labeled and present cited, verifiable facts. Models can choose whether to use them. “This is the opposite of deception,” he said.
The clash highlights deeper tensions. AI companies guard the integrity of their answers fiercely. Once users suspect sponsored content slips in as neutral fact, trust erodes fast. Perplexity itself paused its own ad tests earlier this year. Executives told the Financial Times that labeled ads risked hurting credibility in a market where users can switch to ChatGPT or Claude with one click. The Financial Times reported the decision in February.
Time responded calmly. COO Howard said the publisher contacted Perplexity and looks forward to “a constructive dialogue about the standards and safeguards that will allow advertising and AI to develop responsibly together.” PMI’s CMO, Menaka Gopinath, described the test as a “learning journey” to understand how information on project management, skills, and careers surfaces in AI environments. Not every experiment needs immediate ROI. Sometimes the lessons matter more.
Other major AI players stayed silent. OpenAI and Google declined comment when Business Insider asked. Anthropic did not respond. Their restraint leaves Perplexity as the first to enforce a public policy against these hidden agent ads.
The incident raises questions that stretch beyond one publisher and one startup. The web is splitting. One track serves humans with rich visuals, paywalls, and tracking. The other feeds clean text to agents that summarize, answer, and decide. Publishers see revenue in both. But AI platforms worry that promotional material disguised as authoritative content could pollute training data or outputs. One promotional claim cited as neutral fact could mislead users at scale.
Analysts expect the practice to evolve rather than disappear. Some publishers may block Perplexity’s crawler in retaliation, limiting their visibility in its index. Others will push for industry standards on agent-readable advertising. The debate echoes early search engine optimization wars, where black-hat tactics met algorithm updates and penalties.
Perplexity’s stance carries extra weight because the company positions itself as a truth-seeking alternative to traditional search. Its answers cite sources. It aims for transparency. Allowing cloaked sponsored content would undermine that pitch. At the same time, the firm has faced its own criticism from publishers over content use. The New York Times sent it a cease-and-desist notice in 2024 and later sued over alleged copyright infringement.
Yet here Perplexity acts as gatekeeper. It protects its index from what it sees as manipulation. The message to publishers lands clearly. Serve different content to bots than to humans at your own risk. Downranking follows.
Marketers watch closely. Many experiment with AI visibility strategies. Some optimize websites for crawlers. Others test direct partnerships or paid placements within AI platforms. Time’s test represented a bolder bet: paid influence baked into the data stream agents consume. Perplexity’s block suggests that path faces resistance from the very systems brands hope to reach.
The ad industry debate ignited quickly. Can this format grow into a credible, scalable channel? Or will AI platforms throttle it before it matures? Early evidence points to friction. Perplexity did not wait for broader standards. It acted to safeguard user experience and its own reputation.
Expect more tests. Expect more blocks. The two sides of the web, human and agent, continue to diverge. How advertising fits into that divide will shape revenue models for publishers and trust levels for AI tools. For now, Perplexity has made its position plain. Deceptive tactics, even when labeled sponsored, won’t shape its answers.