The company that started as a nonprofit research lab to build "safe and beneficial" artificial intelligence just told investors its real play is advertising. And the numbers are staggering.

OpenAI is projecting $2.5 billion in advertising revenue this year from ChatGPT, according to Axios, citing a source familiar with recent investor presentations. That is just the warm-up. By 2027, the company expects $11 billion in ad revenue. By 2028, $25 billion. By 2029, $53 billion. And by 2030, $100 billion.

Read those numbers again. OpenAI is telling Wall Street it will be a top-five global advertising company within four years.

The projections rest on an assumption that would have sounded delusional twelve months ago: 2.75 billion weekly ChatGPT users by 2030. That is roughly one-third of everyone on Earth with internet access. For context, Google Search has about 4.3 billion users. Meta has 3.3 billion daily actives across its apps. OpenAI is not targeting a niche. It is targeting the entire planet.

The Numbers Are Working Faster Than Anyone Expected

Here is what makes these projections less ridiculous than they sound: ChatGPT ads are already performing. OpenAI launched its ad pilot in the U.S. in January 2026. Within six weeks, annualized ad revenue crossed $100 million. The company expanded to over 600 advertisers. An OpenAI spokesperson said there was no measurable impact on consumer trust metrics, and ad dismissal rates were low.

That is the nightmare scenario for Google. For twenty years, the entire advertising industry has been built on the assumption that search is the highest-intent advertising surface in existence. Someone searches for "best running shoes" and you show them shoe ads. Simple. Effective. Worth $295 billion a year to Google alone.

But ChatGPT is something search never was: a conversation. When someone asks ChatGPT to help plan a trip, compare insurance providers, or find the best laptop under $1,000, they are revealing intent at a depth Google has never accessed. The ad is not interrupting a search result. It is part of a dialogue. That is a fundamentally different advertising surface, and early data suggests users tolerate it far better than traditional web ads.

The Irony Writes Itself

Remember when Sam Altman said OpenAI would never run ads? Remember when the company explicitly positioned itself as the anti-Google, the company that would put users first and never let commercial incentives corrupt its AI? That era lasted about four years.

Now OpenAI is burning $85 billion a year and has an IPO to justify. The math demands revenue streams beyond subscriptions, and advertising is the obvious lever. But there is a meaningful difference between running a few tasteful ads and telling investors you will be a $100 billion ad company. One is a revenue supplement. The other is a fundamental redefinition of what OpenAI is.

Meanwhile, Anthropic explicitly promised it would never run ads. That positioning looks more strategic by the week. Every dollar OpenAI chases in advertising is a dollar that makes privacy-conscious users reconsider whether the company building their AI assistant has aligned incentives.

The question is not whether OpenAI can build a big ad business. It clearly can. The question is whether you can trust an AI assistant that is being paid to sell you things. First reported by Axios, with additional data from Reuters.