OpenAI rethinks AI shopping strategy after Instant Checkout struggles
OpenAI has scaled back its Instant Checkout feature inside ChatGPT after low adoption and technical challenges, shifting instead toward retailer-built apps and external checkout flows. Reports show users preferred researching products via AI but completing purchases on trusted platforms, leading to poor conversion rates and limited merchant participation. The company is now focusing on improving product discovery and search, positioning ChatGPT as a shopping assistant rather than a full e-commerce platform, reflecting the broader reality that AI-powered shopping is still in its early stages.

Key Highlights
- OpenAI has decided to end its Instant Checkout feature inside ChatGPT after limited success.
- The company will now focus on retailer-built apps within ChatGPT, redirecting users to complete purchases externally.
- Instant Checkout struggled with limited product selection and outdated information.
- AI-powered shopping is still in early stages, with low conversion rates despite growing interest.
- OpenAI is shifting toward product discovery and search, rather than full in-chat transactions.
OpenAI is revising its approach to AI-powered online shopping after its early attempt, called Instant Checkout, failed to deliver the expected results. The feature, launched in 2025, allowed users to purchase products directly inside ChatGPT, but the company is now moving away from that model.
Instead, OpenAI is working with retailers to build dedicated apps within ChatGPT. These apps will guide users through product discovery, but purchases will now be completed on the retailer’s own website or platform.
Why Instant Checkout didn’t work as expected
Despite strong initial interest from major retailers, Instant Checkout faced several challenges. Product selection remained limited even months after launch and the system sometimes showed outdated or inaccurate information such as pricing, stock availability and delivery timelines.
Technical complexity also played a major role. Managing real-time product data across multiple retailers and ensuring accurate transactions proved difficult to scale.
Industry experts noted that enabling smooth in-chat transactions is far more complicated than expected, especially when dealing with inventory, logistics and payment systems across different platforms.
Shift toward apps and better user experience
OpenAI’s new strategy focuses on integrating retailer apps directly into ChatGPT. This approach gives companies more control over customer experience, payments and data, while still allowing users to discover products through AI.
The company is also prioritizing improvements in search and product discovery, where it has seen stronger user engagement.
Rather than acting as a full e-commerce platform, ChatGPT is now positioned more as a shopping assistant that helps users find products and then directs them to complete purchases elsewhere.
AI shopping still in early stages
The shift highlights a broader reality: AI-driven shopping is still evolving. While more users are turning to chatbots for product research and recommendations, most are not yet comfortable completing purchases directly within AI interfaces.
Low conversion rates and user hesitation suggest that adoption will take time, even as tech companies and retailers continue investing heavily in AI-powered commerce tools.
Growing competition in AI commerce
OpenAI’s pivot also comes as competition intensifies. Companies like Google and Amazon are developing their own AI shopping tools, with features such as real-time product updates and deeper integration with retail platforms.
These developments underline how difficult it is to disrupt the e-commerce market, which is still dominated by established players with strong logistics and customer trust.
The bigger picture: AI as a shopping assistant, not a storefront
OpenAI’s decision marks an important shift in strategy. Instead of trying to replace traditional online shopping, the company is focusing on enhancing the discovery phase of the buying journey.
While the vision of fully automated “agentic commerce” is still alive, the current reality shows that AI is better suited for now as a guide that helps users decide what to buy, rather than a platform where transactions happen directly.
