OpenAI is blocking ads inside ChatGPT for image and audio products that compete with its own features, according to reporting from The Information, and the change caught advertisers such as Adobe by surprise. Adobe reportedly lost the ability to promote its Firefly image generator without prior warning, and the restriction was not reflected in OpenAI's published advertising policy at the time it took effect.
The story is small in dollar terms today and large in what it signals. It is an early look at how a company that reaches more than a billion people every week will police a marketplace it both owns and competes in. The specific design of the restriction, not the fact of it, is what makes it worth examining.
What is being blocked, and what is not
The restriction is category-specific rather than blanket. That detail is the whole story.
ChatGPT ad eligibility, by product category
The reported restriction tracks the categories where OpenAI just launched its own tools.
Source: The Information reporting via Search Engine Land, Seeking Alpha and PYMNTS, September 2026. Reported restriction; OpenAI has not published a corresponding policy.
Ads for image-generation and audio-generation tools that compete with OpenAI's own products are reportedly being turned away. Ads for video-generation tools, a category where OpenAI's competitive position is different, can still run. The timing lines up with OpenAI's own launches: the company recently shipped ChatGPT Images 2.5 and a voice product, the exact two categories it is now closing to outside advertisers.
When a platform blocks advertising precisely in the categories where it just launched competing products, and leaves adjacent categories open, the pattern points to a competitive motive rather than a safety or quality rule. That is not an accusation of wrongdoing. It is a description of what the carve-out reveals about intent.
Why this is a textbook self-preferencing question
Economists and regulators have a name for the underlying pattern: self-preferencing. It occurs when a company that runs a marketplace also sells in that marketplace and uses its control of the venue to favor its own products. The scrutiny of Amazon's treatment of third-party sellers and of Google's placement of its own services in search results both turned on versions of this question.
ChatGPT is becoming exactly this kind of venue. As OpenAI turns its user base into an advertising business, it holds two roles at once: the seller of ad space and a competitor to many of the companies that would buy it. Refusing a rival's ad is legal and common in many contexts, and the ad business is young enough that no single decision is decisive. What draws attention is the combination of scale, the two-role structure, and a rule that appears drawn around the company's own product line.
OpenAIThe strategic tension OpenAI is managing
There is a genuine business tension here, and it explains the behavior without needing to assume bad faith.
An advertising business wants maximum demand for its inventory, which argues for accepting every willing advertiser, including competitors. A product business wants to protect its own features from being undercut by rivals advertising to its captive audience. Those two goals point in opposite directions, and OpenAI is currently resolving the conflict in favor of the product.
A platform that both sells ad space and competes for the same customers has to choose which business it is optimizing. The carve-out shows which one is winning right now.
Metir analysis
The cost of that choice is real. Blocking well-funded advertisers such as Adobe removes some of the highest-value demand from a marketplace OpenAI is trying to scale toward a reported multibillion-dollar target. It also risks the perception that ChatGPT is not a neutral surface, which matters because the assistant's value to users rests partly on the belief that its recommendations are not quietly shaped by who OpenAI competes with. A recommendation and an ad are different things, but a user does not always experience them as different, and that is precisely where trust is won or lost.
What to watch next
Three things will tell you whether this is a one-time defensive move or the start of a durable policy.
The first is transparency. Reporting noted the restriction was not in OpenAI's published ad policy. If the rules that govern who can advertise stay unwritten, advertisers cannot plan and regulators will ask why. A clear, published policy is the minimum bar for a marketplace at this scale.
The second is consistency. If the blocked categories keep tracking OpenAI's own product launches, the self-preferencing reading strengthens. If genuine safety or quality criteria emerge that apply evenly to OpenAI and to competitors, it weakens.
The third is the broader lesson for anyone building on top of a single dominant assistant. When one company controls the surface, the model, and the competing products, its commercial incentives can quietly shape what users see. Keeping options open across multiple providers, rather than routing everything through one gatekeeper, is one way to keep that influence in check. OpenAI's ad carve-out is a small, early example of a pattern that tends to grow with a platform's power, and it is worth watching for exactly that reason.
Sources:
- OpenAI reportedly blocks rival AI tools from advertising in ChatGPT | Search Engine Land
- OpenAI limits Adobe and other competitors from advertising in ChatGPT: report | Seeking Alpha
- OpenAI cuts off rivals' ads on ChatGPT | PYMNTS
- OpenAI restricts ads for competing products, impacting Adobe | GuruFocus
Image credits
Pioneer Building, San Francisco, by HaeB, via Wikimedia Commons, licensed under CC BY-SA 4.0.