On February 9, 2026, OpenAI introduced a major shift that the industry had been anticipating for over a year: ads officially went live inside ChatGPT.

For marketers, this represents the early stages of a new advertising environment. A new platform, hundreds of millions of users, and intent signals that make Google Search look like guesswork.

But before your brand starts investing, there’s a conversation the industry isn’t having loudly enough: this isn’t just another ad channel. You’re not buying a banner. You’re not sponsoring a feed. You’re placing your brand inside someone’s thought process. And that changes everything about brand safety, consumer trust, and what it means to show up in the right context.

‍

What’s Actually Happening

Here’s the quick rundown. OpenAI is currently serving ads to users on its free tier and $8/month Go plan in the U.S. If you pay for Plus, Pro, Business, or Enterprise, you won’t see ads. The format right now is contextual text ads placed at the bottom of ChatGPT’s responses, clearly marked as sponsored. OpenAI claims the ads are matched based on conversation topic, past chat history, and previous ad interactions, but that advertisers never see your actual conversations.

On the infrastructure side, there’s no self-serve ad platform yet. Early access appears to be invite-only with reportedly seven-figure minimum commitments. Criteo announced its spot as the first ad-tech partner integrating with OpenAI’s pilot, which signals that programmatic buying may soon follow. OpenAI is using a cost-per-engagement pricing model where advertisers pay when users interact with ad units, not just when they see them.

In the 2024 “A fireside chat with Sam Altman OpenAI CEO at Harvard University,” Altman stated, “Ads plus AI is sort of uniquely unsettling to me…I kind of think of ads as like a last resort for us for a business model.” That shift illustrates how quickly the economics of AI platforms are changing, and why advertising is becoming part of the conversation around long-term sustainability.

Outside of ChatGPT, Anthropic’s Claude has made it clear that they will not be running ads. Demis Hassabis, CEO of Google DeepMind, revealed that Google Gemini doesn’t have any current plans to bring ads to their platform, but seeing as how fast Altman changed direction, that may be up in the air. 

‍

Not Influencer Marketing. Not Search. So What Is It?

Many commentators are comparing ChatGPT ads to Google Search ads or social media placements. But this is a fundamentally different environment, and brands that treat it like another line item in their programmatic budget are going to learn some hard lessons.

When someone searches Google, they’re scanning a list of options. When someone scrolls Instagram, they’re passively consuming content. But when someone is using ChatGPT, they’re often thinking out loud. They’re processing decisions, working through personal problems, researching sensitive topics, or getting help with work they’re accountable for. Anthropic’s own analysis of Claude conversations found that a meaningful portion involves deeply personal or sensitive subjects, the kind of things you’d discuss with a trusted advisor, not a search engine.

That context matters enormously for brand safety. Your ad isn’t appearing next to content like it does on a webpage. It’s appearing inside a conversation, one that feels private, personal, and trusted. If a user is asking ChatGPT about managing anxiety and your meal kit ad shows up at the bottom of that response, the brand association isn’t neutral, it’s intrusive. This is why defining a deeper brand identity has become a requirement for any brand wanting to maintain trust in conversational spaces.

‍

Brand Safety Has a Whole New Definition

Traditional brand safety has always been about adjacency, making sure your ad doesn’t appear next to hate speech, misinformation, or content that clashes with your values. The industry built entire ecosystems around this: keyword blocklists, pre-bid filtering, post-bid auditing.

AI ads introduce a problem those systems were never designed to handle. The content your ad appears alongside is generated in real time by a model that can hallucinate, misrepresent facts, or produce responses with unintended bias. According to an industry analysis, even top-performing models hallucinate between 0.7% and 1.5% of the time. That might sound small, until you consider the volume: hundreds of millions of conversations per week. That margin of error presents a liability for brands running ads on any AI platform. 

‍

Margin of Error

Here’s a scenario brands need to think about: a user asks ChatGPT to compare your product to a competitor’s. The AI generates an answer that’s partially wrong, maybe cites an outdated price, an inaccurate feature, or a capability your product doesn’t have. And right below that flawed response? Your sponsored ad. Now your brand is associated with misinformation about itself. OpenAI vows ads won’t influence responses, but that doesn’t mean the responses themselves will always be accurate. To mitigate this risk, brands must prioritize managing AI citations and accuracy to ensure the models are pulling from a verified source of truth.

Additionally, OpenAI has stated that ads won’t appear near sensitive topics like health, politics, or mental health. But the line between sensitive and non-sensitive in conversational AI is far blurrier than on a static webpage. A conversation can start with recipe research and shift to body image concerns within three messages. The contextual filters that work on web pages were not built for this level of fluidity.

‍

‍

AI Advertising Is Entering a Skeptical Market

Brands considering ChatGPT ads need to understand the room they’re walking into. The general public’s relationship with AI is complicated at best and hostile at worst. According to a 2026 survey of nearly 1,500 Americans, 81% expressed concern about AI accessing their personal data. A separate IAB study found that the gap between how positively advertisers assume consumers feel about AI-generated ads and how consumers actually feel has widened from 32 to 37 points in the last two years, with executives dramatically overestimating public acceptance. Gen Z, ChatGPT’s largest user demographic, is actually growing more skeptical of AI ads, not less.

This isn’t just theoretical backlash. When OpenAI tested app suggestions in late 2025 that resembled ads, users pushed back immediately. Anthropic capitalized on that sentiment with a Super Bowl commercial mocking the concept of AI ads and doubling down by pledging that Claude would remain completely ad-free. It resonated because it tapped into something real: people already feel uneasy about AI, and injecting advertising into that relationship amplifies the discomfort.

‍

The Bigger Picture: AI Trust Is Under Fire

The advertising conversation doesn’t exist in a vacuum. Right now, the broader AI industry is dealing with a credibility problem that should give any brand strategist pause.

For brands, this means a ChatGPT ad doesn’t just carry your reputation, it carries the platform’s reputation too. And that platform’s reputation is being actively debated on the front page.

‍

What Brands Should Be Asking Before They Buy In

None of this means ChatGPT ads are a bad idea. Criteo’s early data shows users referred from AI platforms convert at roughly 1.5x the rate of other referral channels, which makes sense since someone actively problem-solving in a chatbot is high-intent by nature. The opportunity there is real. But the approach needs to be different from anything brands have done before.

How will ads be targeted? OpenAI says targeting is based on conversation topics and past interactions, not keyword bidding. But the infrastructure is still being built. Will targeting eventually get granular enough to match specific prompts? Will advertisers be able to bid on conversational intent the way they bid on search keywords? These details matter enormously, and most of them haven’t been answered yet.

What platforms and tools will manage this? Right now, there’s no public self-serve platform. Criteo’s integration suggests programmatic is coming, but the measurement and attribution tools are immature. Reporting from early 2026 has stated that OpenAI plans for CRM integrations, voice-enabled ads, and dynamic creative optimization later in the year. But those are roadmap items, not current capabilities. Brands buying in today are essentially beta testing.

Will placement be random or contextual? Currently, ads are contextually matched to the conversation topic. OpenAI has said it will show the most relevant advertiser for a given chat. But “relevant” is doing a lot of work here. A conversation about cooking could be relevant to kitchenware, meal delivery, diet programs, or grocery services. Without more granular controls, brands risk showing up in contexts that are technically adjacent but tonally wrong.

What’s the fallback when AI gets it wrong? If ChatGPT hallucinates inaccurate information about your product directly above your sponsored ad, what recourse do you have? This is an entirely new brand safety challenge that no existing verification tool was designed for. The Air Canada chatbot case, where the company was held liable for its bot’s false promise about bereavement fares, should be a cautionary tale for anyone assuming the platform will protect your brand’s integrity.

‍

What Brands Should Take Away

ChatGPT ads represent something entirely new in digital marketing, not just a new channel, but a new category. The rules of search, social, and programmatic don’t fully apply here because the environment is more personal, more fluid, and more opaque than anything brands have advertised through before.

The opportunity is significant. But the brands that win in this space will be the ones who resist the urge to treat it like another line in the media plan, and instead ask harder questions about context, trust, and what it means to show up inside someone’s most private conversations with a machine. Move thoughtfully, not impulsively.