On May 5, 2026, OpenAI did something that would’ve sounded absurd three years ago: it opened a self-serve advertising platform inside ChatGPT. Not a pilot program for Fortune 500 brands with a $50,000 minimum buy-in. A real, functioning Ads Manager where any business — your local coffee shop, a SaaS startup, a dropshipping side hustle — can register, upload creative, set a budget, and start running ads inside the world’s most popular AI assistant.
The target? $2.5 billion in ad revenue this year. And by 2030, OpenAI wants $100 billion from advertising alone.
Let that sink in. We’re watching a company that built its brand on ad-free, subscription-funded AI suddenly bet the farm on the very business model it implicitly defined itself against.
What Actually Launched
The new Ads Manager is a US-only beta for now, but the roadmap is aggressive. OpenAI removed the $50,000 minimum spend that gated its earlier pilot, opening the floodgates to small and medium businesses. Advertisers can now buy on both CPM (cost per thousand impressions) and CPC (cost per click), with conversion tracking via a new Conversions API and pixel-based measurement tools.
The partner ecosystem is already substantial. Agency heavyweights like Dentsu, Omnicom, Publicis, and WPP are plugged in. Ad tech platforms including Adobe, Criteo, Kargo, Pacvue, and StackAdapt can manage campaigns through their existing workflows.
OpenAI’s pitch is straightforward: people use ChatGPT to research products, compare options, and make decisions. Those moments — when someone asks “what’s the best project management software for a 10-person team?” — are commercially valuable. An ad served in that context, theoretically, converts better than a banner on a random news site.
The Revenue Context
This isn’t a side experiment. OpenAI’s U.S. ad pilot reportedly generated over $100 million in annualized revenue within six weeks back in March, according to Reuters. That pace suggests the demand is real and the inventory is undersupplied.
The bigger picture, per eMarketer data cited by Axios, shows AI search advertising growing from roughly $1 billion in 2025 to $25.9 billion by 2029. AI’s share of total search ad spend would jump from 0.7% to 13.6% in just four years. OpenAI is positioning itself to capture a dominant slice of that expansion.
| Metric | 2025 | 2029 |
|---|---|---|
| Total U.S. Search Ad Spend | $148.6 billion | $190.7 billion |
| AI Search Ad Spend | $1.0 billion | $25.9 billion |
| AI Share of Search Revenue | 0.7% | 13.6% |
Source: eMarketer via Axios
Why This Matters Beyond the Balance Sheet
There’s a strategic layer here that gets lost in the revenue headlines. OpenAI’s infrastructure costs are astronomical. Training frontier models and serving hundreds of millions of users doesn’t come cheap. The $20/month ChatGPT Plus subscription was never going to cover those bills at scale. Advertising is the classic playbook for monetizing high-engagement, low-friction consumer platforms — see Google, see Meta, see every social network that ever reached a billion users.
But ChatGPT isn’t a social network. It’s a utility. People trust it precisely because it feels neutral. When you ask for the best hiking boots, you expect a recommendation shaped by quality, not by who paid for placement.
OpenAI knows this is the tension. Their announcement repeatedly emphasized three principles: answers stay independent, conversations stay private, users stay in control. Advertisers get aggregated performance data only. They can’t see what users are discussing. Ads are visually separated from organic answers.
Asad Awan, OpenAI’s head of monetization, told reporters that ads “do not influence the core organic model.” The question is whether users will believe that over time, especially when the same company building the ranking algorithm is also selling the ad slots.
The Privacy Paradox
OpenAI’s privacy safeguards are technically sound on paper. Conversations aren’t shared with advertisers. Personal details are walled off. But the measurement tools — the Conversions API and conversion pixels — still create a data trail. A user clicks an ad in ChatGPT, lands on a website, completes a purchase, and that action is reported back to OpenAI. It’s not personally identifiable in isolation, but it builds a feedback loop that optimizes ad delivery based on behavioral patterns.
This is standard practice in digital advertising. It’s also exactly the kind of surveillance-adjacent infrastructure that privacy advocates have spent years fighting. OpenAI is walking a tightrope: it needs rich measurement data to attract ad dollars, but every pixel placed and every conversion tracked chips away at the “private conversation” brand identity that made ChatGPT appealing in the first place.
Competitive Ripples
Google and Microsoft are watching this closely. Google’s AI Overview ads and Microsoft’s Copilot monetization experiments are already live, but neither has opened a self-serve floodgate at this scale. OpenAI’s move forces the issue. If AI search advertising is going to be a $26 billion market by 2029, everyone needs a self-serve platform, a partner network, and measurement tools that can compete with Meta’s and Google’s existing stacks.
The real threat to incumbents isn’t ChatGPT ads in isolation. It’s the behavioral shift. Every minute someone spends researching inside ChatGPT is a minute they’re not typing into Google. If that research session includes a relevant, well-timed ad that converts, advertisers will follow the attention. And right now, attention is migrating toward AI assistants faster than most brands have adapted.
What Happens Next
I keep coming back to that $100 billion target. It’s not a revenue goal — it’s a statement of ambition. OpenAI isn’t building an ad business to subsidize research. It’s building an ad business that could eventually rival Google’s.
The immediate risk is execution. Can OpenAI scale ad inventory without degrading user experience? Can it prevent low-quality advertisers from flooding the platform in a race to the bottom? Can it maintain the fiction — and maybe it’s not a fiction — that organic answers and paid placements occupy separate, non-influencing spheres?
The longer-term risk is cultural. ChatGPT became a phenomenon partly because it felt like a clean slate. No tracking. No ads. No algorithmic manipulation. Just you and a machine having a conversation. That era is ending. What replaces it will determine whether OpenAI becomes the next Google or something fundamentally different.
For now, the numbers do the talking. $2.5 billion this year. $100 billion by 2030. A self-serve platform open to anyone with a credit card and a product to sell. The ad machine is live. Whether the conversation stays independent — that part is still being written.