For a decade, the digital advertising industry has operated on a two-platform assumption: Google owns search intent, Meta owns social attention. Every media plan in the world has been built around that duopoly.

OpenAI just announced it wants to be a third platform.

In February 2026, OpenAI began rolling out ads inside ChatGPT for users in the United States. The initial rollout targets free and “ChatGPT Go” tier users — Premium subscribers remain ad-free for now. Expansion to the UK, Canada, and Australia is scheduled for Q2 and Q3 2026. For a product used by more than 800 million people every week, this isn’t a pilot. It’s a statement of intent.

The advertising industry should pay attention. Not because ChatGPT ads are going to immediately dethrone Google — they won’t. But because the structural logic of where ads sit inside a ChatGPT conversation is fundamentally different from anything that has existed before. And that difference has real implications for how brands build strategy in 2026 and beyond.

What the ads actually look like

OpenAI is leading with restraint, at least visually. Ads appear at the bottom of ChatGPT’s responses, clearly labelled as “Sponsored Recommendations.” They are contextual — meaning they are matched to the topic of the conversation, not to a user profile built from personal data. OpenAI says conversations remain private from advertisers. They say they do not sell user data.

A few design choices stand out. Ads are blocked from appearing alongside conversations about health, mental health, and politics. They cannot be shown to users under 18. These guardrails are more protective than what either Google or Meta typically applies, and they are clearly designed to preserve what is most valuable about ChatGPT to its users: the sense that it is a trusted interlocutor, not an ad surface.

Whether that trust survives contact with commercial reality is the central question.

The early data, reported by Search Engine Land, is sobering. Click-through rates for ChatGPT ads appear significantly lower than on established search platforms. That was predictable. Users who are mid-conversation — asking a question, processing an answer — are not in the same “browse and click” mindset as a Google Search user who just typed “best running shoes 2026.” The intent is different. The friction is different.

The $200,000 barrier that reveals the real strategy

Here is the detail that most coverage has glossed over: the initial beta programme costs a minimum of $200,000 to enter.

That number is not a media buy. It is a statement about who OpenAI is building this product for. At $200K minimum, you are selecting for enterprise brands, holding companies, and mid-market firms with significant digital budgets. You are explicitly excluding the long tail of small businesses that funds Google and Meta. This is a deliberate starting point, not an oversight.

The strategic logic is clear: OpenAI needs case studies, not volume. They need to be able to show the world’s most prominent CMOs that advertising inside a conversational AI is worth paying for. If the first wave of results is good, the second wave of investment follows — and the price drops.

This is the same playbook Perplexity tried, briefly, before abandoning it. We covered Perplexity’s decision to kill its ads — the company concluded that ads posed too great a risk to the trust its product depends on. OpenAI is betting the opposite: that it can run ads without poisoning the well, because the ads are architecturally separated from the answers.

The Ads Manager that changes the equation

Alongside the ad rollout, OpenAI is testing an Ads Manager dashboard. This is the piece that matters most for marketers who aren’t among the $200K beta participants.

The Ads Manager signals that OpenAI is building the infrastructure for self-serve advertising — meaning, eventually, any brand with a credit card and a campaign brief could buy into ChatGPT the way they buy into Google Ads today. The current beta is the institutional sale. The Ads Manager is the long game.

For agencies, this creates a genuine planning problem. AdWords knowledge became a mandatory skill for every digital team in the 2000s. Facebook Ads Manager did the same in the 2010s. If ChatGPT Ads Manager scales to self-serve — and OpenAI’s $25 billion revenue target for 2028 suggests they need it to — then “ChatGPT campaign management” becomes a skill set that the industry will need to develop, train, and bill for by 2027.

Where this fits in the “Answer Economy”

The biggest structural shift isn’t about ad format — it’s about where purchase decisions are being made.

We have already documented how Google’s AI Overviews are destroying organic click-through rates and how OpenAI’s early CPM data is already changing brand expectations for ChatGPT. The pattern those stories together describe is this: the locus of consumer research is moving from the search results page to the conversational interface.

A user who might have Googled “best CRM software for a 50-person team” in 2023 is now likely asking ChatGPT the same question — and getting a synthesised answer rather than ten blue links. If that ChatGPT answer includes a “Sponsored Recommendation” from a relevant software brand, the brand has placed itself inside the decision-making process at exactly the right moment. That is more valuable than a search ad clicked before the research starts.

This is the “Answer Economy” thesis that a number of marketing commentators have begun building frameworks around. The core idea: as AI models increasingly mediate consumer research, the brands that are recommended by AI — either organically through strong AEO practice or commercially through sponsored slots — are the brands that win consideration.

The complication is the trust premium. Google’s search ads work partly because users know they are ads and have already discounted them accordingly. ChatGPT users are not yet in that habituated relationship with the product. Some proportion of them may not initially realise that sponsored recommendations are paid placements. When they do realise — and they will — ChatGPT’s response needs to demonstrate that the sponsorship hasn’t compromised the answer.

How this should change your planning

Most brands in Q1 2026 cannot afford the $200K entry cost. But that doesn’t mean the ChatGPT ad rollout is irrelevant to their strategy. There are specific actions that make sense right now.

Invest in “recommendability,” not just visibility. ChatGPT’s contextual ads are matched to conversation topic. Organic mentions — the result of a brand being recommended by the model without payment — are matched to the quality of a brand’s information footprint. If your brand’s description on its own website is thin, inconsistent, or jargon-heavy, AI models will either not cite it or cite it inaccurately. Treat your website content as AI training material.

Track your brand’s AI Citation Rate. A handful of tools — including emerging platforms like Profound and Trackta — now measure how often specific brands are mentioned in AI model responses for category-relevant queries. This is the 2026 equivalent of tracking Share of Voice in search. Start measuring it now so you have a baseline before paid ads change the landscape.

Understand the difference between intent layers. ChatGPT advertising will reward brands that meet users at the research and consideration stage. It will likely underperform for brands that need to capture immediate purchase intent — that’s still Google’s territory. Mid-funnel is where the opportunity lives.

Plan for the international rollout timeline. UK, Canada, and Australia come in Q2–Q3 2026. If you are a global brand with budgets that span those markets, your Q2 planning conversations should already include a line item for ChatGPT testing, even at a modest level.

What this means for Google

Google’s response to all of this has been visible but not panicked — at least publicly. The comparison between Google AI Mode and ChatGPT ads shows that Google has been building defensively: AI Overviews, AI Mode, Gemini integrations across the workspace. The company is not ceding conversational territory without a fight.

But the structural challenge for Google is this: ChatGPT is a trusted answer engine that has not yet been associated with advertising. Google has been an advertising platform for 25 years. The moment users start to feel that Google’s AI answers are shaped by commercial relationships — a fair concern, regardless of whether it is technically accurate — ChatGPT’s “we didn’t build this for ads” positioning becomes powerful.

OpenAI is asking the industry to trust that sponsored recommendations and independent answers can coexist. If they deliver on that promise, they will have built something genuinely new. If they don’t — if a year from now the advertising relationship is clearly tilting the quality of responses — users will notice. And they will have done in one year what took Google two decades.

The race for the Answer Economy has started. The $1.3 trillion question is who earns the right to monetise it.