I got served a Facebook ad today that made me do a double take:
Advertise in ChatGPT.
Wait. WHAT?
For years, we’ve been talking about how AI is changing search, how consumers are moving from Google searches to conversational discovery, and how brands need to think beyond traditional SEO.
Now we’ve entered the next phase.
ChatGPT is officially an advertising platform.
And this isn’t a tiny test hiding in a corner of the internet. On August 31, 2026, OpenAI announced that ChatGPT Ads had already reached a $1 billion annualized revenue run rate in less than 200 days, with tens of thousands of advertisers using the platform. ChatGPT Ads are now available in more than 40 countries, with self-service access continuing to expand globally.
For advertisers, this is huge.
But as someone who works in advertising AND uses ChatGPT constantly, my first thought wasn’t actually about CPCs, targeting or campaign strategy.
It was:
Does advertising change the answer?
And maybe more importantly:
Will people still trust ChatGPT if businesses can pay to show up inside it?
First: Where Do the Ads Actually Appear?
This distinction matters.
According to OpenAI, ads appear below ChatGPT’s response, clearly labeled as sponsored content and visually separated from the answer itself.
In other words, if you ask:
“What are the best running shoes for someone training for their first 5K?”
ChatGPT might give you its normal response explaining what to look for and suggesting options.
Then, below that answer, you could see a sponsored ad from a running shoe company.
That advertiser did not pay ChatGPT to become recommendation #1.
OpenAI says advertisers cannot pay to influence, alter, rank or shape the response ChatGPT gives you. The advertising system and the system generating ChatGPT’s answers are separate.
That may sound like a technical distinction, but it is probably the single most important rule OpenAI has put around advertising.
Because otherwise ChatGPT becomes a very different product.
Can Advertisers Pay ChatGPT to Recommend Them?
Currently, no.
OpenAI is explicit about what it calls “answer independence.”
Advertisers cannot pay to change the answer.
They cannot bid their way into ChatGPT’s recommendations.
They cannot tell ChatGPT to rank their product above a competitor.
And an advertisement appearing underneath an answer does not mean ChatGPT or OpenAI is endorsing that company.
Think about the difference between Google’s organic search results and its sponsored search results.
There is an organic information layer.
Then there is an advertising layer.
ChatGPT is attempting to create a similar separation, except the advertising opportunity is based on a conversation rather than a search results page.
And THAT is where things get really interesting for marketers.
ChatGPT Ads Aren’t Really Keyword Advertising
Traditional search advertising starts with keywords.
Someone searches:
“best running shoes for beginners.”
Advertisers bid around those words.
ChatGPT has much richer information about intent.
Someone might instead say:
“I just signed up for my first 5K. I’ve never really been a runner, my knees bother me sometimes, and I’m probably going to train mostly outside. What shoes should I look at?”
That single conversation tells an advertising platform far more than a three-word Google search ever could.
OpenAI says its advertising system looks at things including the context and intent of the current conversation, the advertiser’s landing page, ad copy, creative and advertiser-provided context hints.
Context hints are especially interesting.
Instead of simply telling the platform:
Keyword: running shoes
an advertiser could describe the situations where its product is relevant:
“Cushioned everyday running shoes for beginners training for their first 5K.”
OpenAI says these aren’t exact-match keywords. They are signals that help the system understand the types of conversations, needs and situations where an advertisement could be useful.
This feels much closer to intent targeting than traditional keyword targeting.
And I think we’re going to hear A LOT more about that.
Okay, But Is ChatGPT Reading My Conversations to Serve Me Ads?
This is where things get a little more nuanced.
The ad system can use the context of your current conversation to determine whether an advertisement is relevant.
When personalized advertising is enabled, OpenAI says additional signals may also be used, including past chats, memory and previous interactions with ads.
That sentence is probably going to make some people uncomfortable.
But there is another important distinction:
That information stays within ChatGPT.
OpenAI says advertisers do not receive your conversations, chat history, memories, name, email, precise location or personal details.
Advertisers receive aggregated advertising performance information such as impressions, clicks and conversions. OpenAI also states that it does not sell user data to advertisers.
Users can also turn personalized advertising off and delete their advertising data.
Even with personalization disabled, however, the current conversation may still be used to determine whether an ad is relevant.
So essentially:
ChatGPT can understand what you’re talking about to select an ad. The advertiser doesn’t get to read what you said.
That difference needs to remain crystal clear if OpenAI wants users to continue trusting the platform.
So… Does Advertising Make Me Trust ChatGPT Less?
This is the million-dollar question.
Or apparently the billion-dollar question now.
From a technical standpoint, the answer is not necessarily.
If the advertising system truly remains separated from the answer system, a sponsored placement doesn’t automatically make the information ChatGPT provides less objective.
But perception matters.
Imagine asking ChatGPT:
“Which CRM should I use for my small business?”
ChatGPT gives you an answer.
Immediately underneath it is an advertisement from one of the CRMs you were just discussing.
Even if the answer and advertisement were generated completely independently, some users are naturally going to wonder:
“Did you recommend that because they paid you?”
That is the trust problem OpenAI now has to manage.
And I think the greatest risk isn’t necessarily the advertising product that exists today.
It is maintaining that wall long-term.
The moment advertisers can purchase influence over the actual answer, the value proposition changes dramatically.
For now, OpenAI seems very aware of that risk. Its advertising principles repeatedly emphasize that ads must remain clearly labeled, separate from answers and unable to influence ChatGPT’s responses.
OpenAI also restricts advertising around sensitive conversations where an ad could feel particularly invasive or inappropriate.
Those guardrails matter.
A lot.
What ChatGPT Ads Look Like for Advertisers
Now for the part that made my marketing brain immediately start buzzing.
ChatGPT Ads Manager is becoming a surprisingly familiar advertising platform.
Advertisers can currently build campaigns around objectives including reach, clicks and conversion-focused outcomes. Depending on the objective, campaigns can be purchased using CPM or CPC bidding.
Advertisers can:
- Set budgets and campaign dates
- Target geographic areas
- Target iOS, Android or web
- Create multiple ad groups
- Add conversational context hints
- Upload creative and landing pages
- Track impressions, clicks, CTR, CPC, CPM and conversions
- Add UTM tracking
- Build custom audiences from customer or prospect lists
- Optimize campaigns based on conversion data
OpenAI is even using a relevance-weighted second-price auction, meaning the highest bidder doesn’t automatically win every impression. Relevance is also part of determining which eligible ad gets shown.
That last part is significant.
It suggests that good ChatGPT advertising may depend much more heavily on understanding the problem your customer is trying to solve, not simply having the largest advertising budget.
And Yes, There’s Currently an Ad Credit
OpenAI is clearly trying to get businesses into Ads Manager and testing the platform.
As of August 31, 2026, OpenAI is advertising a promotion for new ChatGPT advertisers:
Spend $500 and receive $500 in ad credit.
Ad credits are subject to eligibility and the specific promotional offer. OpenAI’s general ad-credit terms state that credits normally expire 90 days after issuance unless the offer specifies otherwise.
For advertisers curious about the platform, that makes the barrier to testing it considerably lower.
And yes, I am extremely tempted.
For reference, OpenAI currently lists a $25 daily minimum campaign spend in the United States, and its advertiser documentation recommends starting CPC bids around $3 to $5 per click for click campaigns.
Of course, those benchmarks are brand new and will almost certainly evolve as more advertisers enter the auction.
This Could Be the Beginning of a Completely New Advertising Channel
Google Ads captures someone when they search.
Meta captures someone based largely on interests, behavior, demographics and algorithmic prediction.
ChatGPT advertising potentially captures someone while they are explaining exactly what they need.
That’s a fascinating advertising environment.
Someone isn’t just searching for:
“CRM software.”
They’re explaining:
“I run a five-person company, we’re currently using spreadsheets, I need something easy enough that my sales team will actually use it, and I don’t want to spend $1,000 a month.”
That is extraordinary context.
For marketers, this means messaging may need to evolve from:
What keywords are people searching?
to:
What conversations are people having before they buy?
That is a much bigger shift than simply adding another advertising platform to the media plan.
My Biggest Takeaway
I’m not immediately worried that advertising makes ChatGPT’s answers inherently less trustworthy.
I am going to be watching very closely to make sure the line between advertising and answers stays exactly where OpenAI says it is today.
Because trust is the entire product.
If people begin believing the answer itself is for sale, the usefulness of ChatGPT changes overnight.
But from an advertising perspective?
This is one of the most interesting new channels we’ve seen in years.
We spent the last decade learning how to advertise against searches, demographics, interests and online behavior.
Now we may be entering the era of advertising against intent expressed through conversation.
And that could change a lot more than where we spend our ad budgets.
It could change how businesses think about discovery altogether.
0 Comments