Advertising on ChatGPT has moved from speculation to an active product test. In early February 2026, OpenAI began testing ads inside ChatGPT for logged in adult users in the United States on the Free tier and the Go subscription tier, while keeping paid tiers such as Plus, Pro, Business, Enterprise, and Education free of ads.
For brands and performance teams, that single change opens a new category of inventory: conversational inventory that appears at the moment a user is asking for help, researching options, or making a decision. Unlike most ad environments, ChatGPT is explicitly designed to be trusted for important tasks. OpenAI has repeatedly emphasized that ads do not influence the answers ChatGPT gives, and that conversations are kept private from advertisers.
If you manage growth for a brand, this creates an opportunity and a responsibility. The opportunity is obvious: being present when intent is high. The responsibility is equally important: you will be advertising in a context where users expect usefulness, clarity, and restraint.
What OpenAI is testing today
OpenAI’s current test is designed to be simple and clearly separated from the main ChatGPT answer. In both its product announcement and help documentation, OpenAI describes ads appearing below the end of a response, clearly labeled as sponsored and visually separated from the organic answer.
OpenAI also states that ads run on separate systems from the chat model, and that advertisers cannot shape, rank, or alter ChatGPT’s responses. This matters because it sets expectations for both users and advertisers: you are not buying influence over the assistant’s answer. You are buying a placement that appears after the answer.
There are notable boundaries during the test. OpenAI says it will not show ads in accounts where the user is under 18, and ads are not eligible to appear near sensitive or regulated topics such as health, mental health, or politics.
OpenAI’s help documentation also lists contexts where ads do not appear during this test, including Temporary Chats, when a user is logged out, after generating an image, and in the ChatGPT Atlas browser.
From a marketer’s perspective, these details suggest the product is being introduced with caution, explicit labeling, and strong guardrails. That caution is a signal: OpenAI is optimizing for long run trust more than short run monetization, and it says directly that it does not optimize for time spent in ChatGPT.
The business model behind ChatGPT ads
OpenAI’s rationale is framed around access. In January 2026, OpenAI described ChatGPT Go as a low priced tier and said it planned to test ads to help more people access tools with fewer usage limits or without paying.
That framing matters for advertisers because it influences how the product may evolve. When a platform positions ads as a way to fund broader access, it often means the ad load starts light, the format starts conservative, and the company looks for a balance that preserves user satisfaction. If the test proves that ads can coexist with trust, the inventory could expand. If it harms satisfaction, OpenAI has an incentive to keep ads constrained and rely more heavily on subscriptions.
Independent reporting reinforces that OpenAI is under pressure to fund substantial infrastructure and development costs, which makes monetization experiments like ads more likely to continue in some form.
Why advertising on ChatGPT is different from search and social
At first glance, an ad below an answer resembles a search ad below results. The deeper difference is the interface itself.
Search is keyword first. Social is audience first. ChatGPT is conversation first.
When users interact with ChatGPT, they often provide richer context than a search query: constraints, preferences, timelines, use cases, and follow up questions. That context can signal intent more precisely than many traditional ad inputs, which is why conversational advertising may be especially valuable for complex purchases, higher consideration categories, and services that require explanation.
OpenAI’s own language highlights this direction: the company has suggested that over time ads could become more interactive, where a user can ask questions to make a purchase decision.
For brands, this implies that the creative and landing experience that wins on ChatGPT will not be only about persuasion. It will be about being helpful and immediately relevant to the conversation, because the placement sits right next to an assistant that has already tried to be helpful.
Placement and user expectations: trust is the real auction
In many ad markets, the auction is mainly price and predicted click probability. In a conversational assistant, the real auction includes trust.
OpenAI is attempting to preserve a clear mental model for users: answers are answers, ads are ads, and the two are separated.
That separation does not remove the reputational risk for advertisers. If a sponsored placement feels irrelevant, misleading, or low quality, users may not blame only the platform. They will remember the brand that appeared at the bottom of a sensitive or important conversation, even if the ad was technically eligible.
This is why the early product design matters. A single unit below the answer, shown only when there is a relevant match, suggests OpenAI is limiting frequency and attempting to avoid fatigue.
Targeting and relevance: what signals are used
OpenAI describes ad selection as relevance matching based on the topic of the current chat thread. It also describes the possibility of using additional signals, such as past chats and interactions with ads, when a user opts into personalized ads.
OpenAI also states that it keeps conversations private from advertisers and does not sell user data to advertisers.
Put these together and you get a practical early takeaway: targeting is centered on contextual intent, with personalization controlled on the user side. The most effective advertisers in this environment will not depend on micro targeting. They will depend on building offers and messaging that map cleanly onto common conversation topics.
User controls and what they mean for campaign planning
OpenAI has built multiple forms of user control into the experience. Users can dismiss an ad and provide feedback about why, and OpenAI highlights the ability to learn more about why an ad is being shown.
There is also an option for Free tier users to switch to an ads free experience in exchange for lower usage limits and reduced feature access, with steps in Settings under Ads controls.
Measurement: how to think when the product is still young
OpenAI has not publicly detailed a full measurement suite in the sources above, and it would be a mistake to assume parity with mature ad platforms. The correct approach in an early test environment is to plan measurement in layers.
Layer one is direct response fundamentals: clean conversion tracking on your site, fast mobile pages, and clear event definitions. Even if platform reporting is limited, you can still measure incremental lifts through controlled experiments.
Layer two is intent based learning: treat ChatGPT ads as a source of high intent traffic and evaluate performance by downstream quality, not only by click volume. Many brands will find that fewer clicks can still mean more revenue if the traffic arrives at a moment of strong consideration.
Layer three is qualitative intelligence: what questions did users ask before clicking, and what content did they need next. While you may not have access to private user conversations, you can still learn from aggregate patterns, on site search logs, customer support transcripts, and sales team notes to build better landing journeys.
Creative strategy for advertising on ChatGPT
Because ads appear below an answer and are chosen for relevance, the creative that performs well will likely share a few characteristics.
Be explicit about the use case
Generic brand slogans tend to underperform in intent rich placements. If someone is asking ChatGPT how to choose a mattress for back pain, a mattress brand that speaks directly to support, materials, returns, and delivery will be more credible than a vague promise. You do not need to imitate the assistant’s tone. You need to align with the user’s goal.
Reduce the distance between curiosity and action
A conversational environment encourages exploration. Your landing page should reward that by answering the next three questions a user will ask. If the ad is for a service, show pricing logic, timelines, and examples early. If the ad is for a product, show compatibility, reviews, and clear comparisons early.
Match the post answer mindset
Users will have just received an answer from ChatGPT. That means they are often in evaluation mode, not discovery mode. Ads that work here should feel like a practical next step, not a distraction.
Design for trust signals
OpenAI explicitly warns that seeing an ad does not mean it endorses the advertiser.
That makes trust signals on your own properties more important: transparent policies, clear contact methods, and consistent brand identity.
Brand safety and compliance: do not treat this like any other placement
OpenAI’s early restrictions around sensitive topics and under 18 users show an intent to keep ads away from high risk contexts.
Even with platform level guardrails, brands should add their own.
At minimum, build internal rules about what you will not advertise in conversational environments, and ensure your creative avoids exploiting vulnerability. Categories like finance, health, and personal crises require extra caution, even if OpenAI excludes many of them during the test.
Compliance also includes how you describe your relationship to OpenAI. OpenAI’s brand guidelines and terms emphasize that you must not imply endorsement or misrepresent your relationship with OpenAI, and that use of names and logos must follow their guidelines.
For a company publishing content about advertising on ChatGPT, this is especially relevant: you can talk about the channel, but you should be careful with logos, product marks, and language that implies partnership.
Getting access: what brands can do right now
OpenAI’s help documentation notes that it is focused on learning from early testing and provides a way for businesses interested in advertising to submit their interest to stay informed.
If you are a brand that expects this channel to matter, the immediate objective should be readiness, not rushing spend.
Readiness looks like this:
Strong product market clarity, landing pages built for high intent traffic, reliable analytics, and a creative system that can be iterated quickly.
When access opens more broadly, the brands that have already done that work will learn faster and waste less budget.
A practical roadmap for POMOROI clients
POMOROI’s advantage in emerging channels is not predicting the exact product details. It is building a durable operating system for experimentation and growth. Here is a framework that works well for advertising on ChatGPT specifically.
- First, map the conversations you want to win. Identify the questions customers ask when they are close to purchase. Use sales calls, support logs, reviews, and competitor research to build a list of high intent prompts.
- Second, build landing experiences that answer those prompts. If your page cannot satisfy the next question, you will pay for clicks that do not convert.
- Third, develop creative that is concrete. Focus on outcomes, constraints, and proof. Assume the user is comparing options immediately.
- Fourth, run controlled experiments. Start with limited budgets, tight conversion definitions, and holdout logic where possible. Learn, then scale.
- Fifth, protect the brand. Define exclusions, monitor feedback signals, and ensure messaging remains accurate and respectful in sensitive contexts.
This roadmap is intentionally platform agnostic. It will work whether the test remains small, expands across regions, or evolves into richer formats.
Turn ChatGPT Attention Into Measurable Growth
Advertising on ChatGPT is not just another placement. It is the beginning of conversational advertising as a mainstream channel, introduced with unusually explicit principles around answer independence, privacy, labeling, and user control.
For brands, the winning strategy will not be to force performance tactics into a new interface. It will be to earn attention by being relevant, accurate, and helpful at the exact moment the user is seeking guidance.
If you prepare now, you will be ready to move quickly as the product matures, with a foundation built on trust and measurable outcomes rather than guesswork.
