OpenAI Launches an Advertising Platform: A Rich Person's Business Sold to the Poor

Original Author: Kaori

Original Editor: Sleepy

Sam Altman once called advertising a "last resort" for ChatGPT.

For a long time, that statement represented restraint. OpenAI continued to package itself as a research company, an infrastructure company, a company trying to make AI capabilities accessible to everyone. Advertising, the most familiar monetization method of the old internet, was treated as a backup plan.

But the advertising plan got promoted much faster than expected.

On May 5th, OpenAI launched its self-service advertising platform, Ads Manager, allowing advertisers to place ads on ChatGPT directly or through agencies like Dentsu, Omnicom, Publicis, and WPP. This comes less than three months after the initial ad pilot was launched on February 9th.

The platform is still in a testing phase, but the direction is clear: ChatGPT is no longer just a conversational product; it's also becoming ad inventory. OpenAI aims to achieve $2.5 billion in advertising revenue by 2026 and push ad revenue to $100 billion by 2030.

ChatGPT, with its 900 million user base, is finding the free path increasingly difficult.

Billions in Annual Losses, Relying on Ads to Recover

OpenAI is growing so fast that traditional internet companies can hardly find a benchmark.

But it's also burning through cash quickly.

HSBC analysts estimated in late 2025 that OpenAI could still face a funding gap of $207 billion by 2030. Its cloud and AI infrastructure spending could reach $792 billion between the second half of 2025 and 2030, with long-term computing commitments potentially approaching $1.4 trillion by 2033.

These numbers explain why they are venturing into the advertising business.

Subscription revenue proves users are willing to pay, but it struggles 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 extend the runway, but it dilutes equity and pushes higher valuation pressure back into the company.

Advertising is the fastest source of non-dilutive revenue. It doesn't require free users to pay, doesn't need to re-educate the market, and is easier to pitch to investors.

According to Reuters, OpenAI's ad pilot generated over $100 million in annualized revenue within six weeks. Ads are only shown to free users and Go plan users, do not interfere with ChatGPT's responses, and user data is not shared with marketers.

Setting user privacy aside for now, a more fundamental issue lies behind this strategy.

Ads are sold to free users, but advertisers want to reach paying users. ChatGPT has 900 million weekly active users, with roughly 50 million paid subscribers, translating to a free-to-paid conversion rate of less than 6%. Since ads are only for free users, OpenAI's entire ad inventory comes from that 94% who are unwilling to pay.

The issue is, advertisers willing to spend $50,000 minimum are often not selling products targeting individual consumers. The decision-makers for high-ticket items like enterprise software, SaaS tools, and B2B services are most likely to be ChatGPT's paying users. They spend $20 to $200 a month for better models and larger context windows, and their screens will never see ads.

Beyond the audience mismatch, there's a deeper problem: even if ads successfully reach free users, can the usage scenarios of these users support high ad value?

High Intent Does Not Equal High Conversion

OpenAI's advertising narrative is built on a core assumption: ChatGPT users enter the dialogue with genuine intent, and ad targeting within this high-intent context commands a higher price.

This assumption is only half right.

For the past two decades, brands coveted the search box the most because it represents intent. A user searching for a hotel likely wants to book a room; searching for corporate tax software suggests a potential purchase; searching for the best noise-canceling headphones means the user is already at the decision-making stage.

Google built its advertising empire on this. With ChatGPT, users hand over their decision-making process directly to the AI. This is both more enticing and more daunting for advertisers than search ads. The allure is that ChatGPT sees the entire need; it not only knows what the user wants to buy but also *why*. The daunting part is that if the AI provides the answer directly, the user might never even see the search results page.

But "help me buy a pair of running shoes" and "help me write an email" are two completely different intents. The former is a consumption scenario, the latter a productivity scenario. In ChatGPT's daily usage, the latter accounts for a much larger share than the former. Users come here to write, translate, edit code, draft proposals, and process emotions – high frequency, but not inherently linked to purchasing goods.

This directly suppresses advertising performance metrics. Advertisers are willing to pay a premium for high-certainty purchase intent. Google search ads are expensive because users often enter the search box with clear intentions to buy, compare, book, or order. Meta ads are cheaper, but it possesses social profiles and massive conversion data, using algorithms to repeatedly filter low-intent users into potential consumers.

ChatGPT sits in the middle. It resembles a demand portal more than social media, yet it's harder to gauge commercial intent than search. It's more private than search, yet harder to attribute conversions. It can solve user problems, but it doesn't necessarily generate ad clicks.

This is why OpenAI's shift from CPM to CPC isn't just a product upgrade – it's because advertisers are unwilling to pay based on the "next-gen search engine" narrative for the long term. They ultimately want to know: who generated this click? Where did the conversion happen? How much budget should be moved from Google, Meta, and TikTok to ChatGPT?

Category fit is also a problem. Low-risk categories like home goods, travel, education, and software tools can experiment first. But high-margin categories are often highly regulated, such as finance, healthcare, insurance, and recruitment. If ChatGPT runs ads in these areas, the platform must bear not just ad performance risks but also risks of misinformation, discrimination, and non-compliance.

Google serves as a mirror. In Q1 2026, Google's search ad revenue was $77.25 billion. Yet even Google is very cautious with ad placement in AI Mode and AI Overviews, and its standalone Gemini app still hasn't formally introduced ads.

OpenAI's expansion into advertising is essentially exploring a broader business model for the entire large model track.

OpenAI must make users feel the AI is intimate enough while convincing advertisers that there's enough commercial intent. If this balance is lost, ChatGPT could lose both sides: users might find it inauthentic, and advertisers might find it unconvertible.

But the change brought by advertising doesn't stop there; it's also reshaping how brands compete.

The Focus of GEO is Shifting

Over the past year, brands have worried about disappearing from AI responses. The market packaged this as GEO (Generative Engine Optimization), but it's essentially just old search marketing anxiety repackaged for the AI era.

OpenAI launching Ads Manager hits this anxiety right on the head, but it also shifts the direction of that anxiety.

In the ad-free era, the core question of GEO was "how to get into the AI's context." Brands tried to be cited by the model through product documentation, media reports, third-party reviews, and community discussions. It was a competition of information quality and data structure.

With the ad platform online, targeted traffic can be purchased directly, and brands don't have to rely solely on organic citations. However, the competitive focus hasn't just reverted to "buying more exposure." Instead, it has shifted from "how to get into the AI's answer" to "what does the AI evaluate about my product?"

The reason is simple: after seeing an ad, the user's most natural next step is to ask the AI, "Is this product actually good?" The AI's answer then becomes the true gatekeeper of conversion. Advertisers can buy exposure, but they cannot buy a positive review from the AI. If the AI, based on public data, gives a negative review, every dollar spent on ads accelerates user churn instead of driving conversion.

This means brands need to build a positive reputation within the AI's evaluation system. Product quality itself, the density of user reviews, the coverage of third-party evaluations – these signals the AI can read – will determine conversion effectiveness more than the ad placement itself.

GEO is moving from "entering the context" to "winning the evaluation," a key trend worth watching following OpenAI's new ad platform launch.

Not Advertising is the Most Expensive Ad in 2026

Speaking of OpenAI, we must mention its archrival Anthropic, which is pursuing a completely different "advertising model."

On February 4, 2026, two days before the Super Bowl, Anthropic published a blog post stating that Claude will never run ads. No sponsored links, no third-party integrations.

That statement itself is an expensive ad.

Super Bowl ads aren't cheap. Anthropic spent heavily to tell users it doesn't sell ads, essentially using advertising to buy a brand identity associated with being ad-free.

Being ad-free is never just a moral stance; it's also a business positioning. It tells enterprise clients, professional users, and users in sensitive scenarios that Claude's responses won't be influenced by advertisers, Claude's product direction won't be optimized around ad inventory, and Claude's revenue comes directly from your payment.

The effect was immediate. Claude's ranking on the US App Store climbed from 42nd place at the beginning of the year. On February 28th, after OpenAI's Pentagon contract sparked the "QuitGPT" movement, Claude topped the US App Store's free app chart for the first time ever, surpassing ChatGPT. Free active users grew 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, with annualized recurring revenue soaring from roughly $9 billion to $19 billion. Enterprise tools like Claude Code and Cowork have already contributed at least $1 billion in revenue. Anthropic doesn't need the ad value of free users; it needs the trust premium from enterprise clients who are confident their data won't be used for advertising.

In this context, "no ads" is a precise business decision: using the forgoing of ad revenue to fortify the trust barrier with enterprise clients, thereby supporting higher subscription pricing.

However, "no ads" is not an eternal virtue.

Stanford AI Index data shows that the cost of achieving GPT-3.5-level performance dropped 280-fold in two years, from $20 per million tokens in November 2022 to $0.07 in October 2024. If model capabilities continue to converge and API price wars intensify, the enterprise subscription premium Anthropic enjoys today could be gradually eroded. When model costs drop to the point where all competitors offer similar performance, why would enterprise clients continue to pay more for Claude?

This question currently has no answer, but time will tell.

There's No Such Thing as a Free Lunch

OpenAI chooses advertising; Anthropic chooses to turn ad-free status into a premium. They seem like opposite paths, but both answer the same fundamental question: When the inference cost of an AI product cannot be sustained by a freemium model in the long run, who pays?

OpenAI's Ads Manager is more than just an ad product; it's a signal that the AI industry is moving from free expansion into cost recovery.

But the method OpenAI chose to stop the bleeding reveals the most fragile part of this business model. It needs its user base with the least consumer intent to support an ad pricing model three times more expensive than Meta's.

This is not a problem that can be solved by user scale alone. 900 million weekly active users is an impressive number, but if those 900 million come to ChatGPT to write emails instead of shopping, advertisers will ultimately vote with their feet.

Advertising can be a revenue source for AI products, but it shouldn't be treated as the only answer. Because when a product's business model requires users to stay as long as possible and reveal as much intent as possible, that product ceases to be the user's assistant. It becomes the advertiser's assistant.