Sam Altman once called advertising ChatGPT's "last resort." For a long time, this was a restrained statement. OpenAI still presented itself as a research company, an infrastructure company, and a company trying to make AI capabilities accessible to everyone. Advertising, the most familiar monetization method of the old internet, was treated as an alternative. But the advertising solution came quickly. On May 5th, OpenAI launched its self-service advertising platform, Ads Manager, allowing advertisers to place ads directly or through agencies such as Dentsu, Omnicom, Publicis, and WPP on ChatGPT. This was less than three months after the initial advertising pilot on February 9th. The platform is still in the testing phase, but its direction is clear: ChatGPT is no longer just a conversational product, but is also beginning to become an advertising inventory. OpenAI's goal is to achieve $2.5 billion in advertising revenue by 2026 and push it to $100 billion by 2030. With 900 million users, ChatGPT has found that the free model is becoming increasingly difficult. OpenAI is losing billions annually, relying on advertising to recoup its losses. OpenAI is growing rapidly, so rapidly that traditional internet companies have few comparable benchmarks. However, it's also burning through cash quickly. HSBC analysts estimated at the end of 2025 that OpenAI could still face a $207 billion funding gap by 2030. Its cloud and AI infrastructure spending could reach $792 billion between the second half of 2025 and 2030, and its long-term computing power commitment could approach $1.4 trillion by 2033. These figures explain why they are investing in advertising. Subscription revenue can prove users are willing to pay, but it's difficult to cover the inference costs of all free users. Enterprise APIs can contribute cash flow, but face price wars and model convergence. Capital financing can sustain operations, but it dilutes equity and pushes higher valuation pressures back into the company. Advertising is the fastest non-dilutive revenue source. It doesn't require free users to pay, doesn't require re-educating the market, and is easier to explain to investors. According to Reuters, OpenAI's advertising pilot generated over $100 million in annualized revenue within six weeks. Ads are only available to free and Go plan users, do not affect ChatGPT's response generation, and do not share user data with marketers. Leaving aside user privacy, this strategy masks a more fundamental problem: Ads are sold to free users, but advertisers are looking for paying users. ChatGPT has 900 million weekly active users, with approximately 50 million paid subscribers, resulting in a free-to-paid conversion rate of less than 6%. By only offering ads to free users, OpenAI's ad inventory comes entirely from the 94% who are unwilling to pay. The problem is that advertisers who can afford to spend at least $50,000 are often not selling products aimed at individual consumers. The decision-makers for high-priced categories like enterprise software, SaaS tools, and B2B services are precisely the ones most likely to be ChatGPT's paying users. They spend $20 to $200 per month on more powerful models and larger context windows, while ads never appear on their screens. Beyond audience mismatch, there's a deeper issue: even if ads successfully reach free users, how much advertising value can these users' usage scenarios actually support? High intent does not equal high conversion. OpenAI's advertising narrative is built on a core assumption: ChatGPT users enter the chat with genuine intent; ad reach in this high-intent scenario is worth a higher price. This assumption is only half right. For the past two decades, brands have most wanted to dominate the search box because it represents intent. When a user searches for hotels, it means they might want to book a room; when they search for corporate tax software, it means they might want to make a purchase; when they search for the best noise-canceling headphones, it means the user is already at the threshold of a consumption decision. Google built its advertising empire on this. With the advent of ChatGPT, users directly handed the decision-making process over to AI. This is more tempting, but also more terrifying, for advertisers than search ads. The tempting aspect is that ChatGPT sees a complete demand; it not only knows what the user wants to buy, but also why they want to buy it that way. The terrifying aspect is that if AI directly provides the answer, the user might not even look at the search results page. However, "Buy me a pair of running shoes" and "Write me an email" are two completely different intentions. The former is a consumption scenario, the latter a productivity scenario. In ChatGPT's daily use, the latter accounts for a much larger proportion than the former. Users come here to write, translate, modify code, create plans, and process their emotions—high-frequency, but not naturally linked to product purchases. This directly lowers advertising performance metrics. Advertisers are willing to pay high prices for high-certainty purchase intent. Google search ads are expensive because users often enter the search box with a clear intent to buy, compare, book, or place an order. Meta ads are cheaper, but they have social profiles and massive conversion data, which can use algorithms to repeatedly filter low-intent users into potential consumers. ChatGPT is caught in the middle. It's more like a demand entry point than social media, but it's harder to determine commercial intent than search. It's more private than search, but harder to attribute. It can solve user problems, but it doesn't necessarily generate ad clicks. This is why OpenAI's shift from CPM (cost per mille, pay per impression) to CPC (cost per click) isn't just a product upgrade; advertisers are unwilling to pay long-term based on the vision of a "next-generation search engine." Ultimately, they need to ask: Who brought this click? Where did the conversion occur? How much of the budget should be diverted from Google, Meta, and TikTok to ChatGPT? Category compatibility is also an issue. Low-risk categories like home furnishings, travel, education, and software tools can be tested first, while high-profit categories are often heavily regulated, such as finance, healthcare, insurance, and recruitment. If ChatGPT advertises in these areas, the platform will bear not only the advertising effectiveness but also risks of misleading, discrimination, and compliance. Google's approach serves as a mirror. In the first quarter of 2026, Google's search advertising revenue reached $77.25 billion. Even so, Google remains very cautious about ad placement in AI Mode and AI Overviews, and the standalone Gemini app has yet to officially feature ads. OpenAI's expansion into advertising is an exploration of broader business models for the entire large-scale AI ecosystem. OpenAI needs to make users feel that AI is approachable enough, while also convincing advertisers that there is sufficient commercial intent. If this balance is lost, ChatGPT will lose both sides: users will feel it's not pure, and advertisers will feel it can't convert. But the changes brought about by advertising don't stop there; it's reshaping how brands compete. The focus of GEO is shifting. Over the past year, brands have been anxious about whether they will disappear from AI responses. The market has packaged this as GEO, but it's not essentially a new concept; it's just an old search marketing anxiety repackaged in the AI era. OpenAI's launch of Ads Manager addresses this anxiety, but also shifts its focus. In the ad-free era, the core issue for GEOs (Generational Experts) was "how to enter the AI context." Brands competed for model citations through product documentation, media reports, third-party reviews, and community discussions, focusing on information quality and data structuring. With the ad platform, targeted traffic can be purchased directly, and brands no longer rely solely on organic referrals. However, the competitive focus hasn't shifted back to the traditional "buying more exposure," but rather from "how to get AI's answer" to "how does AI evaluate my product?" The reason is simple: after seeing an ad, the most natural next step for users is to ask AI, "Is this product actually good?" AI's answer then becomes the true conversion gateway. Advertisers can buy exposure, but they can't buy positive reviews from AI. If AI gives negative reviews based on publicly available data, every penny spent on advertising accelerates user churn instead of driving conversions. This means brands need to build a positive reputation within AI's evaluation system. The quality of the product itself, the density of user reviews, and the coverage of third-party evaluations—these signals that AI can glean are more decisive for conversion results than the advertising itself. GEO's shift from "entering context" to "winning reviews" is a trend worth noting after OpenAI launched its new advertising platform.
Not running ads is the most expensive ad of 2026
Speaking of OpenAI, we must mention its arch-rival Anthropic, which is taking a completely different "advertising model."
On February 4, 2026, two days before the Super Bowl, Anthropic posted a blog stating that Claude will never run ads. No sponsored links, no third-party integrations.
This statement itself is an expensive advertisement.
Super Bowl ads aren't cheap. Anthropic spends heavily to tell users they don't sell ads, essentially buying brand awareness of an ad-free environment through advertising. Ad-free isn't just a moral stance; it's also a business positioning. It tells enterprise clients, professional users, and users in highly sensitive scenarios that Claude's answers won't be influenced by advertisers, Claude's product direction won't revolve around optimizing ad inventory, and Claude's revenue comes from what you pay for. The effects were immediate. Claude's ranking on the US App Store climbed steadily from 42nd at the beginning of the year. On February 28th, after OpenAI signed a Pentagon contract, sparking the QuitGPT movement, Claude topped the US App Store's free app chart for the first time, surpassing ChatGPT for the first time ever. Free active users increased by 60%, daily registrations quadrupled, and paid users doubled within a week. Anthropic's revenue structure is completely different from OpenAI's: over 80% comes from enterprise clients, and annualized recurring revenue has surged from approximately $9 billion to $19 billion. Enterprise tools like Claude Code and Cowork have contributed at least $1 billion in revenue. Anthropic doesn't need the advertising value of free users; it needs the trust premium from enterprise clients that their data isn't being used for advertising. In this context, not advertising is a precise business decision. By forgoing ad revenue, it strengthens the trust barrier with enterprise clients, thereby supporting higher subscription pricing. However, "not advertising" is not a timeless virtue. Data from the Stanford AI Index shows that the cost of achieving GPT-3.5 equivalent performance decreased 280 times in two years, from $20 per million tokens in November 2022 to $0.07 in October 2024. If model capabilities continue to converge and an API price war breaks out, the enterprise subscription premium that Anthropic enjoys today may be gradually eroded. When model costs drop to a point where all competitors can offer similar performance, why should enterprise customers continue to pay more for Claude? This question remains unanswered, but time will tell. There's no such thing as a free lunch. OpenAI chose advertising, while Anthropic chose to turn not advertising into a premium. These seem like two opposite paths, but they both answer the same question: when the inference costs of AI products can no longer be covered by a free model in the long run, who will foot the bill? OpenAI's Ads Manager is not just an advertising product; it's also a signal that the AI industry is moving from free expansion to cost recovery. However, OpenAI's chosen method of stemming the bleeding precisely exposes the most vulnerable aspect of this business. It requires using a user group with the least intention to purchase to support an advertising price three times higher than Meta. This is not a problem that can be solved by user scale. 900 million weekly active users is an impressive number, but if these 900 million people come to ChatGPT to write emails rather than buy things, advertisers will eventually vote with their feet. Advertising can be a source of revenue for AI products, but it shouldn't be considered the only answer. Because when a product's business model requires users to stay as long as possible and expose their intentions as much as possible, that product is no longer a user's assistant; it's an advertiser's assistant.