The ads arrived quietly on a Sunday night.
When OpenAI flipped the switch February 9, it marked the end of an era for ChatGPT—the one where the world's most popular AI chatbot existed, at least for its free users, in a commercial vacuum. Now, tucked below the bot's answers on certain queries, a new label appears: "Sponsored."
The company had telegraphed the move for months. There was the December delay, blamed on quality concerns. Then came CFO Sarah Friar's defense at Davos in January, where she called advertising "essential to create a strong business model" that could broaden access. But the real tell arrived this week in reporting by Adweek: OpenAI is requiring brands to commit at least $200,000 to participate in the beta test. Some potential advertisers, according to ad-tech firm Adthena, have been quoted as much as $250,000.
That price tag isn't just a barrier to entry. It's a declaration. OpenAI isn't building a self-serve ad platform where local boutiques and DTC startups can bid a few dollars per click. It's positioning ChatGPT as premium inventory, the kind of real estate that only major brands—working through holding companies like WPP, Dentsu, and Omnicom—can afford to occupy.
A Careful Rollout, By Design
For now, the test is deliberately narrow. Only logged-in U.S. users on ChatGPT's free tier and the new $8-per-month Go plan will see ads. That means the company's roughly 800 million weekly users worldwide—a figure that still lands somewhere between staggering and hard to verify—remain mostly untouched.
Higher-paying subscribers on Plus, Pro, Business, Enterprise, and Education tiers won't see sponsored content. Neither will anyone under 18, or users engaging in temporary chats, generating images, or navigating sensitive topics like health, mental health, or politics. Logged-out sessions stay clean, too.
The mechanics are straightforward, at least as OpenAI describes them. Ask ChatGPT for recipe ideas, and a meal kit or grocery delivery ad might appear below the response—visually separated, clearly labeled, and theoretically relevant to what you just discussed. The company insists the ad doesn't interfere with the answer itself. Conversations remain "model-driven and separate from advertiser influence," according to OpenAI's blog post announcing the test.
Users can interact with an ad via a three-dot menu that lets them "Ask ChatGPT" about the product, but even then, the company claims advertisers have no sway over those replies. Whether users trust that firewall is another question entirely.
What Advertisers Get—and Don't Get
OpenAI's privacy pitch is as much about reassurance as it is about mechanics. Brands receive aggregate metrics: views, clicks, nothing more granular. No chat histories. No stored memories. No personal details. "We never sell your data to advertisers," the company states flatly in its help documentation.
Targeting relies primarily on the current chat thread. If you've enabled personalization—and most users probably have, whether they realize it or not—ChatGPT may also reference prior conversations, stored memories, and previous ad interactions. General location and language factor in. What OpenAI hasn't mentioned is whether external web browsing activity feeds the system, a notable omission given how other platforms weave browsing data into their targeting engines.
Users can dismiss an ad, provide feedback, or delete their ad data "with one tap," as the company puts it. Free-tier users also have the option to go ad-free, though that choice comes with stricter daily message limits—a nudge, perhaps more than a genuine alternative. Upgrading to Plus or Pro removes ads entirely without sacrificing access, which is likely the path OpenAI hopes most engaged users will take.
Who's Already In
Adobe emerged as an early partner, testing ads for Acrobat Studio and Firefly through a deal executed with WPP. In a blog post, the company quoted Asad Awan, OpenAI's head of ads and monetization, who reiterated the separation between organic answers and sponsored content. It's the kind of language that's become standard in digital advertising, though the conversational nature of ChatGPT adds a wrinkle—ads that appear mid-dialogue feel different than banners on a search results page.
Omnicom Media Group disclosed that more than 30 clients have secured placements, spanning apparel, automotive, beauty, consumer packaged goods, hospitality, retail, quick-service restaurants, tech, and telecom. That breadth suggests appetite across verticals, even with the test still invite-only and geographically constrained.
According to Digiday, OpenAI's outreach has been selective—working through major holding companies rather than pitching brands directly. That approach keeps the rollout manageable and gives the company time to refine mechanics before scaling. It also means smaller brands, no matter how flush with venture capital, are locked out for now.
Industry chatter on Reddit and elsewhere has surfaced rumors of $60 CPMs and impression-based billing, though OpenAI hasn't officially confirmed those figures. The company maintains an interest sign-up page for advertisers but hasn't published self-serve specs, API documentation, or finalized pricing models beyond the $200,000 minimum.
The Competitive Landscape
OpenAI isn't breaking new ground here—Google has been slipping ads into AI Overviews throughout 2025, and Microsoft's Copilot includes Sponsored Links through its Chat Ads API, which also powers ads inside Snapchat's My AI chatbot. Both companies emphasize separation and clear labeling, much as OpenAI does now.
Perplexity tried a similar experiment starting in late 2024 but hit a wall. TechCrunch reported last October that the company paused accepting new advertisers, citing gaps in measurement and tooling. The pause was revealing: monetizing conversational AI at scale isn't just a question of inserting ads—it's an operational challenge that requires infrastructure most platforms don't yet have.
Then there's Anthropic, which has positioned Claude as the ad-free alternative. The company even ran a Super Bowl spot mocking ChatGPT's ad rollout, a move OpenAI criticized as misleading. The exchange felt less like a substantive debate and more like a preview of how divergent monetization strategies will play out among leading AI labs. One company's business model is another's brand liability.
Regulatory Clouds on the Horizon
U.S. Senator Ed Markey isn't waiting to see how this plays out. In January, he sent letters to OpenAI and other companies, flagging concerns about privacy, manipulation, and youth safety in ad-supported chatbots. Forbes reported that Markey urged caution on conversational ad models—a signal that regulatory scrutiny may intensify as more platforms monetize AI interactions.
OpenAI has excluded users under 18 and blocked ads in sensitive contexts, but those guardrails may not satisfy lawmakers looking for more robust protections. Job listings suggest the company is building an ads-integrity team to handle brand safety and restricted verticals beyond the initial exclusions, though finalized policies remain unpublished.
What Comes Next
Several questions hang in the air. When will OpenAI expand ads beyond the U.S.? Will the ad-free Free option persist long-term, or will it quietly disappear once the company has enough paying subscribers? And what about self-serve buying tools and third-party verification—the kind of infrastructure that makes digital advertising legible to brands outside the top 100?
The test is just that, at least officially. But with a $200,000 floor, major holding-company partners already committed, and a privacy narrative carefully constructed, OpenAI is treating ads as a core revenue stream, not a side experiment. The company's bet is that conversational AI is valuable enough, and ChatGPT's user base large enough, to command premium pricing even as the mechanics remain uncertain.
Whether that bet pays off depends on data the U.S. pilot will deliver in the coming months. And whether the rest of the industry—Google, Microsoft, Perplexity, and whoever else joins the fray—can keep pace with the challenge of monetizing billions of AI-driven conversations without breaking the experience that made those conversations valuable in the first place.
For now, the ads are live. And OpenAI, after years of speculation about its business model, has finally shown its hand.
