Why Are There Ads in ChatGPT Now?
For most of the internet's history, advertising has tried to answer one valuable question:
What does this person want?
Google found an extraordinarily profitable answer.
Look at what someone searches for.
A person typing "best running shoes" into Google is probably thinking about running shoes. A company selling them may be willing to pay for the chance to appear at exactly that moment.
That simple idea helped turn search advertising into one of the largest businesses on the internet.
ChatGPT creates a different possibility.
A person might begin with:
"I'm thinking about running my first 10K."
There is no shopping request yet.
The conversation continues. They run three times a week, mostly on pavement. Their current shoes become uncomfortable on longer runs. They have a budget.
Eventually, they ask which shoes might suit them.
A conversation that started with running has become a purchase decision.
That difference may explain why advertising inside AI could become much more important than a few sponsored boxes appearing underneath answers.
Google learned how much a search could be worth.
Now the AI industry is beginning to learn something else:
What is a conversation worth?
Why Put Ads in ChatGPT at All?
The simplest answer is that AI is expensive.
Every question sent to a large AI system requires computing infrastructure. Serving hundreds of millions of people means paying for data centers, chips, electricity, networking and enormous amounts of computing capacity.
Subscriptions are one way to pay for it.
Advertising is another.
OpenAI began testing ads in ChatGPT in the United States in early 2026. Ads are shown on Free and Go plans, while paid tiers such as Plus and Pro remain ad-free.
The ads are labeled as sponsored and displayed separately from ChatGPT's answers.
Free internet services supported by advertising are nothing new.
But the place where these ads appear is unusual.
They appear inside a conversation.
And the business is already becoming significant.
In August 2026, less than a year after the first tests, Reuters reported that ChatGPT Ads had reached a $1 billion annualized revenue run rate.
The experiment was becoming a business.
A Search and a Conversation Reveal Different Things
Imagine two people looking for a hotel.
The first searches Google:
"best hotel Tokyo"
That reveals quite a lot.
The person probably wants a hotel in Tokyo.
Now imagine someone tells an AI:
"I'm going to Tokyo with my six-year-old son. I want somewhere quiet, but I still want to walk to restaurants. We'll stay four nights and I don't want to spend more than $300 a night."
Both people may eventually book a hotel.
But the second interaction contains more context.
The AI knows the destination, budget, length of stay and what kind of neighborhood matters to the traveler.
This is one of the fundamental differences between search and conversational AI.
Search often captures a request.
Conversation can reveal how a decision is being formed.
For advertisers, that difference could be valuable.
ChatGPT Can Use the Context of the Conversation
OpenAI says its advertising system can consider the context and intent of the current conversation when deciding which advertisement may be relevant.
That creates a different kind of advertising opportunity.
Someone discussing photography might eventually need a camera.
Someone planning a vacation might eventually need a hotel.
Someone starting a business might eventually need software.
The commercial opportunity does not always begin with a commercial question.
It can emerge during the conversation.
The valuable advertising space, then, may not simply be a keyword.
It may be the moment when a conversation becomes a decision.
But Advertisers Do Not Get Your Conversation
There is an important distinction here.
Using conversational context to select an advertisement is not the same thing as giving the conversation to an advertiser.
OpenAI says advertisers do not receive users' chats, chat histories, memories or personal details.
It also says user data is not sold to advertisers.
Advertisers can receive aggregated performance information, such as how many times an advertisement was viewed or clicked, without receiving the conversation itself.
So it would be misleading to describe the business as:
"OpenAI sells your conversations."
A better way to think about it is this:
The conversation isn't being sold. The opportunity to appear at a relevant moment is.
That distinction is important because it explains what the advertising inventory actually is.
Google Already Proved the Value of Intent
Advertising becomes more valuable when it appears close to a decision.
A television commercial can reach millions of people who have no intention of buying a car.
A Google search for:
"BMW X3 lease price"
is different.
The audience may be smaller, but the commercial intent is much stronger.
That is one reason search advertising became such a powerful business.
Conversational AI could move one step earlier in the process.
A user may not begin by looking for a product.
They may simply be trying to solve a problem.
But solving a problem often leads to a purchase.
Which software should I use?
Where should I stay?
Which laptop fits my work?
AI assistants increasingly participate in the thinking that happens before those decisions.
That position could be commercially valuable without replacing search itself.
The Advertisement Cannot Become the Answer
There is an obvious problem.
Imagine asking an AI:
"Which laptop should I buy?"
Now imagine the company paying the most for advertising could secretly influence the answer.
The usefulness of the assistant would immediately become questionable.
Search engines have dealt with a version of this problem for decades by separating paid advertisements from organic results.
Conversational AI faces an even more sensitive version of it.
People don't only use AI to find websites.
They ask it to explain, compare, recommend and reason.
OpenAI says ads are separate from the systems generating ChatGPT's answers and do not influence those answers. Sponsored placements are also labeled separately.
This separation is not merely a product-design decision.
It is part of the economics.
An AI assistant has something valuable that ordinary advertising inventory does not automatically have:
trust in the answer.
If advertising damages that trust, the advertising space itself becomes less valuable.
The product is therefore not simply attention.
It is attention without destroying the credibility of the assistant that attracted the attention in the first place.
From Portals to Search to Feeds to Conversations
Internet advertising has repeatedly changed when a new place for human attention appeared.
Early internet portals sold space around content.
Google turned search intent into advertising inventory.
Facebook and Instagram monetized feeds.
YouTube monetized video attention.
Amazon monetized product searches.
Conversational AI introduces another possible surface.
But it is not simply another rectangle on a screen.
A user can ask a question, add information, compare alternatives, change their mind and eventually make a decision without leaving the same conversation.
That makes the decision process itself potentially valuable advertising territory.
The opportunity is large enough that the industry is already debating where the boundary should be.
When OpenAI first announced its advertising plans in January 2026, the Associated Press highlighted the tension between monetizing a huge free audience and preserving users' trust in a tool they may use for personal questions and advice.
A month later, Reuters described the same challenge from the business side: advertising can generate enormous revenue, but building a mature ad platform requires measurement, infrastructure and, above all, user trust.
Those problems do not disappear simply because the product uses AI.
They may become more important.
Google Monetized the Search. AI May Monetize the Conversation.
The first era of internet advertising was largely about attention.
Put an advertisement where lots of people are looking.
Search improved that model by adding intent.
Put an advertisement where someone is actively looking for something.
Conversational AI could add another layer:
context.
Not just what someone typed into a search box, but what problem they are trying to solve and where they are in the process of solving it.
That is why a small sponsored placement underneath an AI response could represent something much larger.
The interesting question isn't simply why ChatGPT has ads now.
Lots of internet products have ads.
The interesting question is what kind of advertising market can exist when software participates in the conversation that happens before a purchase decision is made.
Google built an enormous business around knowing what people search for.
AI may create another valuable moment:
the moment someone figures out what they actually want.
BEYOND THE OBVIOUS.
Sources
OpenAI — Our Approach to Advertising and Expanding Access
OpenAI — Information for Advertisers
Reuters — OpenAI's Ad Business Hits $1 Billion Annualized Revenue Run Rate
Associated Press — ChatGPT's Free Ride Is Ending: OpenAI Plans Advertising on the Chatbot