Advertising in AI-native destinations and experiences is already starting to look less like a series of experiments and more like an emerging market.
Three months ago, I argued that AI platforms had started acting like media companies. Now, the conversation has already changed. It wasn’t simply that OpenAI introduced ads in ChatGPT and advertisers started to experiment. It was that an entire ecosystem of buying tools, measurement, agency partnerships and commerce infrastructure began forming around AI media.
Here are the three biggest developments in Q2 2026 and what I’m watching in Q3.
#1: OpenAI had an advertising coming-out party
OpenAI got off to a fast start with advertising in ChatGPT in Q2; after launching in February 2026, ads are already available in eight countries around the world. The company spent most of Q2 developing the tools and features advertisers expect from a digital ad business, including a self-serve ad manager, conversions API and CPC bidding.
“We are clearly in the advertising business now.” — Denise Dresser, chief revenue officer, OpenAI, at Cannes Lions
What types of companies were ChatGPT’s biggest advertisers during the quarter? Retailers, according to Sensor Tower data I wrote about in this newsletter in early June.
However, these businesses didn’t advertise themselves; instead, they placed reams of product ads. AI shopping conversations naturally revolve around products rather than stores, suggesting AI media may reward product relevance over retailer brand awareness.
Another important development ad industry executives should pay attention to: OpenAI not only wants to sell ads on ChatGPT; it also wants advertisers and their agencies to use its creative tools to make and manage them. During a Cannes Lions press briefing that I attended, the company showed a demo of how agencies could use Codex and OpenAI’s workspace agents to develop creative assets and variations.
Essential reading
[OpenAI’s Dave Dugan explains how ChatGPT is building a new ad category](https://www.businessinsider.com/openais-dave-dugan-explains-the-strategy-behind-chatgpt-ads-2026-6)
[OpenAI moves to automate ad creative](https://digiday.com/marketing/openai-moves-to-automate-ad-creative/)
[83% of ChatGPT ad triggers don’t exist in traditional search](https://www.pymnts.com/news/artificial-intelligence/2026/83percent-chatgpt-ad-triggers-dont-exist-in-traditional-search/)
Takeaway
Q2 2026 was an inflection point for OpenAI. Now it’s time for the company to prove it can earn more than just experimental ad budgets. That will happen if it creates ad formats that are genuinely unique, delivers better measurement and attribution than its competitors and becomes an end-to-end resource for both the creative tools to make ads and the media channels to place them.
#2: The AI media infrastructure race accelerated
Every successful advertising medium has the same supporting infrastructure: Measurement. Buying. Optimization. Attribution. Creative. Commerce connections. AI media made rapid progress toward checking off almost every one of those boxes in Q2.
What’s most interesting is how quickly adtech, creative tech and ad agency partners have moved to partner with OpenAI.
The company created partnerships with Criteo, StackAdapt, Pacvue, Kargo and Adobe. Holding companies Dentsu, WPP, Publicis and Omnicom joined as agency partners, and LiveRamp came on board for measurement (showing that first-party data and identity are becoming foundational to AI-era marketing).
Outside of OpenAI, a few other developments point to the intersection of commerce infrastructure with AI media. The FIDO Alliance, an organization founded to create interoperable password authentication standards, started to work on standards for agentic commerce, alongside Google and Mastercard. And Albertsons and Criteo are putting sponsored products into conversational grocery search.
Essential reading
[OpenAI Opens ChatGPT Ads to Self-Service Platform](https://www.adweek.com/media/openai-opens-chatgpt-ads-to-self-service-platform/)
[OpenAI’s ChatGPT ads get their first measurement partner in LiveRamp](https://digiday.com/marketing/openais-chatgpt-ads-get-their-first-measurement-partner-in-liveramp/)
[Google donates Agent Payments Protocol to FIDO Alliance](https://blog.google/products-and-platforms/platforms/google-pay/agent-payments-protocol-fido-alliance/)
[Brands Can Now Buy Albertsons Conversational Search Ads Through Criteo](https://www.adweek.com/commerce/exclusive-brands-can-now-buy-albertsons-conversational-search-ads-through-criteo/)
Takeaway
One clear sign of a maturing ad market is the availability of infrastructure that makes it easier to plan, buy and optimize ads and then measure performance and compare results with other types of advertising. Even while advertisers continue to test AI media, the supporting infrastructure is making those experiments easier to execute. Advertisers don’t need to shift major budgets overnight, but they do need to start building institutional knowledge about how advertising in conversational settings performs differently than other forms of advertising.
#3: Google delivered a coherent vision for how advertising will work in AI
Before Q2, Google’s AI initiatives—and the advertising within them—felt disconnected. By the end of the quarter, however, Google connected the dots. It delivered a coherent narrative for how search and AI will coexist and how advertising will work in AI-powered search, shopping and video.
At Google I/O, the company announced the integration of AI tools directly into the search box, calling it “the biggest upgrade to our Search box in over 25 years.” It also plans to introduce information agents that automate basic search tasks and agentic booking capabilities in search.
At Google Marketing Live, the company delivered new ad formats in AI Mode including conversational discovery ads and highlighted answers, and said it plans to introduce AI-powered shopping ads and business agents that exist directly within an ad. In addition, the Direct Offers test is being expanded to travel advertisers, with new features including the ability to bundle promotions and the integration of native checkout for Universal Commerce Protocol merchants.
*Google’s Highlighted Answers ad format *
With these moves, Google clarified how search evolves in tandem with AI, not separate from it, and how the company’s sprawling ecosystem of ad products is starting to evolve toward a conversational future.
Essential reading
Google announces new AI Mode ad formats and agentic commerce tools at Google Marketing Live
[A new generation of ads for the AI era of search](https://blog.google/products/ads-commerce/google-marketing-live-search-ads/)
[Google pushes AI-generated ads further into search results](https://www.wsj.com/cmo-today/google-pushes-ai-generated-ads-further-into-search-results-6b478d83) (paywall)
Takeaway
Google is actively redefining what search means—for consumers as well as advertisers. It’s becoming an AI assistant that does the work for people: answering questions, comparing the products, monitoring news, and soon, completing purchases. That’s teeing up a big change in how search advertising looks and feels.
What to Watch in Q3 2026
**AI ad formats start to become more unique. **Google’s upcoming new ad products tell me that we’re about to enter a phase when the ad formats in AI media really start to come into their own. Not a replica of search, but truly unique formats that not only exist in AI destinations but use AI as part of the ad experience itself. Recent job listings on OpenAI’s jobs site offer important clues about new image, video and interactive ad products that might be coming.
New ways to attribute conversions. One thorny question in the ad industry is the ability to measure the effect of ads in AI media. What do consumers do next? Do they drive conversions or sales? In Q3 I’m expecting OpenAI, Google and others will push to deliver better attribution tools to help advertisers understand the AI customer journey and the proximity of AI platforms to the decision click.
**Will advertisers stick with ChatGPT after initial tests? **Growth has already been rapid—more than 1,400 brands have advertised on ChatGPT in the US in the first couple of months, across nearly every industry category, per Sensor Tower. But I’m less interested in the number of new advertisers on the platform (which will naturally go up given its expansion to new markets and the addition of self-serve) and more interested in the percentage of advertisers that stick around (or come back after they evaluate their initial results).
Retailers and AI media get closer together. Amazon and Walmart, the two largest players in retail media, are already taking steps to integrate their AI search tools (Alexa and Sparky, respectively) more deeply into consumers’ everyday shopping behavior. Neither has said much about the role of paid advertising in these AI tools, but I expect that will change. Conversational ads in retail media will enable retailers to capture consumers who are becoming accustomed to searching for and finding products by using natural language queries, not keywords.
**Meta AI stumbles (again). **Anyone familiar with my past work at eMarketer knows I spent nearly two decades researching social media ad businesses. If you had asked me two years ago whether Meta would be a leader in AI media, I would have said, “Absolutely!” I assumed it would surely figure out how to extend ads into AI-native ad surfaces. But that hasn’t happened. While it’s using AI to make its existing ad surfaces and formats perform better, it hasn’t been able to come up with AI features that consumers want and that don’t get the company in trouble. I don’t expect Meta to materially change that picture in Q3.
The Bottom Line
-
Ads in AI media are no longer a curiosity; they’re becoming increasingly accessible to advertisers.
-
The supporting infrastructure to optimize, measure and compare performance is arriving faster than expected.
-
Now is the time to build institutional knowledge, because the market is maturing faster than most planning cycles.
This column originally appeared in Debra Aho Williamson’s weekly Substack The AI Ad Economy.
“Data-Driven Thinking” is written by members of the media community and contains fresh ideas on the digital revolution in media.
*Follow Debra Aho Williamson and AdExchanger on *
LinkedIn.
For more articles featuring Debra Aho Williamson, click here.