In February 2024, Sam Altman told an interviewer that advertising was a "last resort" and that "ads-plus-AI is sort of uniquely unsettling to me." At Harvard, he called it a "momentary industry."
On February 9, 2026, OpenAI launched advertising inside ChatGPT.
The shift from public position to product launch took eighteen months. What followed in the weeks after the launch tells a more detailed story about where AI advertising currently stands, how quickly it is evolving, and what the early data actually shows.
How ChatGPT Advertising Currently Works?
Ads appear at the bottom of ChatGPT responses, visually separated from the answer and labelled sponsored. Users on the free tier and the $8 per month Go plan see ads. Paid subscribers on Plus, Pro, Business, Enterprise, and Education plans do not.
OpenAI launched with a CPM model at $60 per thousand impressions and a minimum spend of $200,000 to $250,000. Within ten weeks, the CPM had dropped to as low as $25. On April 15, OpenAI switched to a cost-per-click model with bids between $3 and $5 per click, minimum spend reduced to $50,000, and a self-serve ads manager opened to global advertisers.
Targeting is contextual, matched to the topic of the current conversation rather than to demographic or third-party data. Advertisers cannot see user conversations, chat history, names, email addresses, or IP addresses. They receive aggregated performance data showing total views and clicks.
Early advertisers in the pilot included Target, Ford, Adobe, Mrs. Meyer's, and Expedia. Within two months of launch, the pilot had crossed $100 million in annualised revenue. OpenAI is projecting $2.5 billion in advertising revenue for 2026, scaling to $11 billion by 2027 and $100 billion by 2030.
What Makes This Surface Different From Search and Social?
The advertising surfaces that have defined digital marketing for the past two decades share a structural characteristic: the user arrives with a defined intent or a passive browsing state. Search captures users mid-query. Social intercepts them mid-scroll.
ChatGPT operates differently. Users go there to reason through decisions, to compare options, evaluate tradeoffs, and work out what they actually think before committing to a course of action. Research conducted by ad tech firm Adthena and cited by Digiday describes this as intent building through the back-and-forth of prompt-driven conversations, a different quality of engagement than either search or social has historically captured.
This distinction matters for how advertising performs in the environment. A click on a ChatGPT ad, in theory, comes from a user who has already formed a question and is actively working through the answer. Whether that translates to better conversion data is something the industry is still measuring. OpenAI's measurement infrastructure, including attribution models, incrementality testing, and media mix modelling, is still being built. CNBC reported in March 2026 that early results were moving too slowly to meet the hype and that OpenAI cannot yet prove the ads are working due to the absence of mature measurement tools.
The Broader Shift: Every Major AI Platform Now Has Advertising
ChatGPT's advertising launch was the highest-profile but not the only significant development in AI advertising in the first half of 2026.
Google began integrating ads into AI Overviews, the AI-generated summaries that now appear at the top of search results for most informational queries. Research from Seer Interactive tracking 3,119 queries across 42 organisations found that organic click-through rates fell 61% on queries where AI Overviews appear. Paid click-through rates fell 68%. Brands cited inside AI Overviews, however, saw 35% higher organic CTR and 91% higher paid CTR than non-cited brands on the same queries, a finding that has significant implications for how brands approach content strategy.
Microsoft launched AI Max for Search in open pilot in May, applying expanded query matching and personalised assets across Copilot and Bing. The announcement was framed around winning across all three eras of the web: the human web, the LLM web, and the emerging agentic web where AI agents make purchases on behalf of users.
Amazon launched Sponsored Products Prompts and Sponsored Brands Prompts inside Rufus, its AI shopping assistant, in March. Rufus handles an estimated 274 million daily queries. Amazon's internal data shows sessions involving Rufus are twice as likely to result in a purchase.
The Trust Data
The advertising launches have occurred against a backdrop of measurable user scepticism.
An Ipsos survey found that nearly two-thirds of US adults say ads in AI search make them trust the results less. A campaign called QuitGPT gathered more than 200,000 sign-ups protesting the introduction of ads into ChatGPT.
Anthropic's response to the OpenAI launch was direct. The company spent millions running Super Bowl commercials with titles including Deception, Betrayal, Treachery, and Violation, with the tagline Ads are coming to AI. But not to Claude. The spots resulted in an 11% jump in daily active users for Claude within days of airing.
Perplexity, which had tested advertising in 2024 and 2025, abandoned it entirely. The company is now targeting $500 million in subscription revenue and has positioned itself explicitly as an ad-free AI search alternative.
The divergence between platforms is significant. OpenAI and Google are building advertising into their AI products. Anthropic and Perplexity are building subscription models around the explicit absence of advertising. Which approach users prefer and which generates more durable revenue remains an open question.
Where the Industry Stands
AI advertising is real, live, and generating revenue. The infrastructure is early. The measurement tools are still being built. The user response is mixed and being watched closely by every platform, every advertiser, and every researcher in the space.
What is established: every major platform where buyers now conduct research and form preferences has introduced an advertising surface in the first half of 2026. The contextual targeting model, matching ads to conversation topic rather than user demographics, is a structural departure from how digital advertising has worked for the past two decades.
What is not yet established: whether the contextual targeting is precise enough to deliver the conversion rates that would justify significant budget reallocation from search and social. Whether users will habituate to ads in AI conversations the way they habituated to ads in search results. Whether the trust deficit measured in the Ipsos survey represents a temporary friction or a durable ceiling.
