ChatGPT is starting to test ads in conversations: here’s what’s really changing

What Changes And What Stays The Same

OpenAI has begun testing ads in ChatGPT for some users in the United States, adding a new revenue stream while stating that answers will remain neutral and conversations will stay private from advertisers. The experiment focuses on free and lower-cost tiers and is designed to expand access to advanced AI without paywalls for every user.

Who Will See Ads In ChatGPT?

OpenAI is rolling out ads gradually, starting with logged‑in adult users in the US on the Free and ChatGPT Go plans. Higher‑tier subscriptions, including Plus, Pro, Business, Enterprise, and Education, remain ad‑free for now. The test began in early 2026 and will expand over the following weeks as OpenAI collects feedback and adjusts the experience. The company’s stated goal is to learn how ads can fit “naturally” into ChatGPT without undermining user trust.

Plans, pricing and ads

For now, upgrading from Free or Go to a higher tier is the simplest way to avoid ads if you want an entirely ad‑free **experience**.

How And Where Ads Will Appear

OpenAI is testing clearly separated, labeled ad units instead of weaving promotions directly into ChatGPT’s written answers. Ads are visually distinct, marked as “Sponsored” or similar, and displayed in blocks or banners that are separate from the main response, typically at the bottom or side of the conversation window. They are designed to feel contextual to what you are discussing, so they may appear next to a response rather than appearing as part of the generated text itself. OpenAI’s principle is that ads should be useful and context‑appropriate, not jarring interruptions that distract from the answer.

Will Ads Influence ChatGPT’s Answers?

OpenAI explicitly states that ads do not influence how ChatGPT generates its responses. Chat answers are produced on infrastructure that is separate from the ad system, and advertisers cannot change, rank, or bias what the model says. Sponsored placements are not endorsements, so the presence of an ad does not mean OpenAI recommends that product or service. The company frames ads as a way to fund broader access to powerful AI while preserving trust in the assistant’s neutrality, meaning that in theory you should receive the same answer with or without an ad alongside it.

How Ad Personalisation Works

Ad targeting in ChatGPT focuses on conversational context and limited behavioral signals rather than classic keyword search advertising or broad third‑party tracking. During the test, OpenAI may use the topic of your current chat, patterns across past conversations, and your prior interactions with ads (such as clicks or dismissals) to select which sponsored message to show.

For example, if you ask for recipes or meal‑planning ideas, you might see ads for meal kits or grocery delivery services that align with that theme. When ad personalization is turned on, historical interest signals can be layered on top of real‑time context so that advertising stays relevant to both your long‑term interests and the immediate conversation.

What advertisers can and cannot see

OpenAI says advertisers do not get access to your individual chats, memories, or identifiable profile data. Instead, they receive only aggregated metrics, such as total impressions and total clicks, without user‑by‑user breakdowns. The company commits that it will not sell user data to advertisers, addressing a major concern of surveillance‑style targeting. Conversation content is processed by the ad system for topic classification and safety, but transcripts themselves are not shared with brands or ad buyers.

User Controls, Privacy And Data Protection

OpenAI is pairing the ad rollout with a set of privacy controls and transparency tools aimed at giving users more say over personalization. You will be able to view your ad interaction history, clear it at any time, dismiss specific ads, and provide feedback about whether an ad was useful or inappropriate.

“About this ad”‑style explanations will show why a particular ad appeared, referring to factors like conversation topic or high‑level interest categories. You can also manage personalization preferences so that less of your usage data is used for targeting while still keeping normal ChatGPT features enabled.

On the privacy side, OpenAI stresses that advertisers cannot see your conversations, stored memories, or personal identifiers such as your name or email address. Reporting to advertisers is aggregated, and the company reiterates that it does not sell user data.

Ads are required to be clearly labeled and separated from answers, and OpenAI says the model is not tuned to favor advertisers’ products in its reasoning or recommendations. Temporary chats, where available, remain ad‑free and do not contribute to ad personalization, offering a more private option for sensitive queries.

Public Reaction And Rival Responses

The move toward an ad‑supported model has sparked debate about whether conversational AI should mix deeply personal interactions with commercial messaging. Privacy advocates worry that even contextual ads around chats about health, relationships, or mental well‑being could feel intrusive, even if underlying data is not directly shared with advertisers.

Industry analysts see the test as a signal that large‑scale AI services may need some form of advertising or cross‑subsidy if they are to stay widely accessible at low or zero cost.

Anthropic’s Super Bowl campaign

Anthropic, maker of the Claude assistant, ran high‑profile Super Bowl commercials that dramatized scenarios where people seeking personal or emotional advice are steered toward awkward commercial offers. Each spot ends with the tagline “Ads are coming to AI. But not to Claude,” clearly positioning Anthropic against ad‑supported chatbots without naming ChatGPT directly.

The company followed up with a public statement promising that Claude will remain ad‑free, arguing that advertising inside conversations conflicts with its vision of a deeply helpful assistant.

Sam Altman’s response

OpenAI CEO Sam Altman pushed back on X and in interviews, describing Anthropic’s campaign as misleading and “obviously dishonest” in how it portrays hypothetical AI ads. He insisted that OpenAI would “never run ads in the way Anthropic depicts them” and emphasized that users would reject those kinds of intrusive experiences.

Altman also criticized Anthropic for using what he called “deceptive advertising” to attack ad formats that do not exist in ChatGPT’s current test. The exchange highlights a broader strategic split: OpenAI is embracing ads as a revenue stream, while Anthropic is using an ad‑free Claude as a differentiator.

What This Means For Everyday Users

For most people on the Free or Go plans in the US, ChatGPT will start to resemble other free online services where contextual advertising helps pay for infrastructure and new features.

You will occasionally see sponsored blocks around your chats, tailored to what you are talking about, but the underlying answers are meant to remain the same whether an ad is shown or not. You retain control over ad personalization, can review and clear your ad history, and can adjust how much of your usage data is considered for targeting.

If you prefer an ad‑free experience, higher‑tier plans like Plus, Pro, Business, Enterprise, and Education remain available, and alternative assistants such as Claude have publicly committed to staying free of ads within the chatbot itself.