ChatGPT Ads Have Arrived: What's Next & How Brands Can Prepare

ChatGPT ads are here. Sam Altman once called advertising a "last resort"; well, now that last resort has arrived. Brands are already lining up for the beta, scrambling to figure out what this means for their 2026 marketing plans.

I'll cut to my take: this is a cautious experiment, and most brands won't see meaningful value from ChatGPT ads this year. It's also one of the riskiest moves OpenAI could make. But the broader story isn't about whether this particular launch succeeds, it's about what happens when AI becomes an advertising surface.

The Promise (At Least on Paper)

Let's be honest about why this is exciting, even if it's dangerous. OpenAI is sitting on something no advertising platform has ever had: a live stream of intent-rich conversations powered by a highly capable reasoning engine. When someone asks ChatGPT to help plan a trip, compare software tools, or research a major purchase, they're revealing far more about what they want and why than they ever would in a search query.

If ads could tap into that context responsibly, this might be the strongest intent signal advertising has ever seen. The amount of personalization ChatGPT could infer makes Google's keyword targeting look primitive by comparison.

Now combine that with agentic commerce, where ChatGPT doesn't just recommend products, but actually completes transactions; add a commission-based model where OpenAI takes a cut, and the potential revenue machine is obvious.

Why This Is Incredibly Hard

All of this being said, running ads on top of an intelligence layer is fundamentally different from running ads in search results or social feeds. And the challenges here are substantial.

The Trust Problem is Existential

An assistant that knows intimate details about my life, my preferences, and my anxieties, and then nudges me to buy something that doesn't clearly align with my interests will feel manipulative. It doesn't matter how many guardrails you promise.

The relationship between user and AI assistant is different from the relationship between user and search engine. We tolerate ads in search because we understand the transaction: Google gives us answers, we accept some commercial noise. But we're starting to think of AI assistants as something closer to advisors. Advisors who take money to steer your decisions lose your trust fast.

The Media Narrative Will Be Brutal

OpenAI is volunteering to be the punching bag of the AI industry. Edge-case horror stories will surface quickly: the AI that recommended an expensive product to someone in debt, the assistant that pushed a brand deal when a user was vulnerable. I can already see the headlines in The Verge and The New York Times. These stories will shape public perception far more than any carefully worded blog post from OpenAI.

The Competitive Timing Couldn’t Be Worse

This experiment is happening while Gemini gains consumer traction and Anthropic wins enterprises. The performance gap between frontier models is narrowing. Claude has explicitly said it won't run ads. Gemini doesn't run ads in the consumer experience today. Competitors will weaponize this moment to sharpen their positioning around trust.

The Inventory Math Doesn’t Work Yet

Early access appears limited to logged-in U.S. adults with constrained placements; ads appear below the answer, not within it. That's not enough impression volume to support a healthy auction. Without scale, without competitive bidding, the economics stay experimental.

The Bigger Picture: AI as an Advertising Surface

Zoom out from ChatGPT specifically, and you’ll find that something important is happening: AI search has become a mainstream channel. ChatGPT, Gemini, Perplexity, AI Overviews, and AI Mode aren't curiosities anymore. They're among the most effective channels influencing buyers, often ahead of traditional search and social when you measure downstream impact.

Our data shows that ChatGPT-sourced leads convert at nearly 2x the rate of Google search leads. That's not a rounding error. That's a signal that the funnel is collapsing: people are discovering products and making decisions inside these AI interfaces, not just using them as a starting point.

So, even if ChatGPT's first ad product underwhelms, the direction is clear. Both Google and OpenAI will pursue AI advertising revenue despite skepticism. The question isn't whether AI ads are coming; it's what shape they take.

What Brands Should Do Now

1. Fix Your Measurement Stack First

AI ads won't show value with last-click attribution. If your measurement framework can't handle top-of-funnel influence and multi-touch attribution, you'll either over-invest or under-invest. Get incrementality measurement in place before you spend a dollar.

2. AEO Becomes Essential

Here's the key detail: OpenAI has said the ad doesn't influence the actual answer. The ad sits below the response. That means visibility inside of the answer itself (the citations, the recommendations, the organic response) still matters most. Answer Engine Optimization isn't optional anymore; it's how you show up in the part users actually trust.

3. Expect More Tooling to Follow

Ads bring infrastructure. Expect an "AI ad manager," ChatGPT Analytics, and more granular performance data. OpenAI will give advertisers access to outcomes and reporting even if the underlying auction logic stays opaque. Start thinking about how AI surfaces fit into your media mix now, not after the tooling matures.

4. Build Your Control Plane Across AI Surfaces

Between ChatGPT, Gemini, Perplexity, AI Overview, and whatever comes next, fragmentation is accelerating. Brands need systems to manage visibility and performance across these surfaces, not just one-off experiments.

Where This Might Actually Work

If I had to bet on a model that succeeds, it's not impression-based ads below the answer. It's agentic commerce with a commission structure.

Imagine ChatGPT helps you plan a trip, books your flights and hotels, and OpenAI takes a small cut from each transaction. In that model, OpenAI stays at an equal distance from all merchants; there's no incentive to push a worse product because the commission is the same. The AI remains your ally because its interests align with yours: to complete a transaction that you're happy with.

That's a harder technical and business problem than slapping ads below responses. But it's also the version that might actually preserve user trust while building a durable revenue stream.

ChatGPT Ads: The Real Question

My guess is that this initial push is as much about proving a revenue story supporting fundraising and investor confidence as it is about winning the ad market in the near term. OpenAI needs to show that it can monetize beyond subscriptions, and ads are the obvious lever.

But the fundamental tension remains: the more ChatGPT feels like a product designed to sell you things, the less it feels like an assistant you can trust. And trust is the entire value proposition.

We're about to watch OpenAI try to thread that needle. The next few months will tell us whether it's possible (or whether AI advertising requires a completely different model than what we've built for search and social).

Either way, the AI ads era has arrived. It's going to be messy. And every brand needs to start paying attention.