OpenAI held its annual developer day on Monday. The company introduced its plan to build apps directly into ChatGPT. The demonstration was impressive, showing how programs like Spotify and Figma can be accessed without ever leaving the ChatGPT window. With the tech world rapidly moving towards AI integration, OpenAI’s demo provided the clearest picture yet of what an AI-first internet could look like, where interfaces like ChatGPT can query information and execute commands directly.
If you were watching closely, you may have noticed there is significant potential for financial transactions within this system. Just last week, the company launched Instant Checkout, an agentic shopping system that serves as payment infrastructure for one-off purchases. This system connects to any stores that sell through Shopify, Etsy, or Stripe. Now, the new apps feature provides the front-end infrastructure, allowing service providers to build their own interfaces directly into ChatGPT.
In short, OpenAI now has all the necessary components for AI-driven commerce, establishing ChatGPT as a destination where customers go to buy and retailers go to sell. This represents a huge new line of business for the company, one with major implications for the entire tech industry. In this new model, OpenAI is not just competing with Google and Anthropic, but also with giants like Amazon and Walmart.
A look at OpenAI’s announced app partners reveals the ambitious scope of this vision. ChatGPT will be able to call you a cab through Uber, book a trip on Expedia, call a plumber or locksmith through Thumbtack, order groceries from Instacart, prepared food from DoorDash, or big-box goods from Target. Without much more development, ChatGPT could become a portal for most of its users’ discretionary spending.
If this system works, it would be worth far more than a twenty-dollar monthly subscription. The precise financial terms are still unclear, but like any app store, OpenAI is well positioned to take a portion of all money spent on its platform. Furthermore, ChatGPT will recommend products, leveraging its vast data about users, which tips the balance of power even further in OpenAI’s favor. In this scenario, ChatGPT becomes a super-aggregator, funneling customers to retailers and serving as an entry point for ever-larger amounts of commerce. While OpenAI has many potential business directions to pursue, it is no exaggeration to say that AI-driven commerce is one of its most lucrative options.
OpenAI is not the only company focused on this opportunity. On the same day as the ChatGPT announcement, Adobe released a report predicting that this year’s holiday season would be dominated by AI-assisted shopping. The report suggests shoppers will turn to chatbots instead of search engines to find the best deals. A separate report from Mastercard identified agentic commerce as a new competitive arena for finance. Google has already launched its own competing protocol for agentic commerce called AP2. This protocol has a broader scope but less current momentum than OpenAI’s version.
The simplest form of AI-driven shopping is using ChatGPT to find products instead of a search engine. For example, if you are looking for a canvas tennis shoe under eighty dollars, ChatGPT can find it for you just as easily as Google Search. However, AI systems do not need to be passive. The AP2 specification includes a provision for agent-initiated purchases. This could allow an agent to buy concert tickets the moment they become available or book a flight as soon as the price drops below a certain point. Of course, there could also be agents on the other side of the transaction, negotiating with purchasing agents for the best deal and bundling goods under the right circumstances. If retailers and customers are willing to take the leap, the changes could extend far beyond a simple buy button.
The biggest unanswered question is whether the shopping public will actually be interested. AI shopping is obviously attractive to OpenAI, and companies like Stripe and Mastercard see plenty of benefit in it too. However, users have not shown much interest in agentic shopping systems beyond simple product searches. But then again, they have not had a real chance to use them. These systems are not even properly available now, and it will be months before the average user can try out a fully agentic shopping system. When they finally do, there will be a lot riding on how they react.