OpenAI’s AI now picks the ad format: ChatGPT carousels arrive 5 months after ads debut

The launch of product carousels in ChatGPT marks more than an incremental ad format update; it’s a demonstration of how large language models are being wired into programmatic advertising. Here, an AI not only serves an ad but decides its structural form, pulling from external data feeds and optimizing for engagement in real time. This blurs the line between conversational assistant and automated marketplace, raising new technical and ethical questions for AI-driven monetization.

OpenAI has taken a decisive step forward in monetizing ChatGPT with the introduction of product carousel ads, a format that moves beyond the simple single-product placements that have been live for approximately five months. Verified by Digiday via screenshots, the new format allows multiple products from a single retailer to appear side-by-side within a dedicated ad slot at the bottom of a user’s conversation. This evolution is not just cosmetic; it alters the dynamic between advertiser and platform, shifts ChatGPT closer to the transactional shopping experiences offered by search and social media, and sets the stage for a more competitive and algorithmically driven ad ecosystem. The carousel arrives just three months after OpenAI enabled retailers to bulk-create "product feed" campaigns — a feature that automated the ingestion of entire product catalogs rather than requiring manual, one-off ad builds. That earlier update, rolled out around May 2026, was a clear signal that OpenAI was building infrastructure for richer, more scalable e-commerce advertising. The product carousel is its first public realization.

“

OpenAI has taken a decisive step forward in monetizing ChatGPT with the introduction of product carousel ads, a format that moves beyond the simple single-product placements that have been live for approximately five months.

The mechanics of the carousel borrow heavily from Google Shopping’s feed-based approach. OpenAi pulls product data directly from retailers’ feeds and uses its own internal systems to determine when to show a carousel versus a single ad, and which specific products to feature. Today, all items in a carousel come from the same retailer, but the architecture supports a more disruptive future: multiple retailers appearing side by side, with products ranked by relevance, intent matching, or bidding factors. This would mirror the familiar, high-stakes competition that already defines product listing ads on Google. For any advertiser accustomed to granular control over ad formats, the shift may be jarring — OpenAI’s black-box decisioning is not only choosing which ads to serve, but also whether to pack multiple products into one placement.

For the broader advertising industry, the development has significant implications. First, it validates conversational AI interfaces as viable ad real estate beyond simple search. ChatGPT’s ad products are evolving along a trajectory that directly parallels the early days of social media and search monetization, where formats grew from static images and text links to dynamic, feed-based, and shoppable units. Second, the automation of ad format selection mirrors the trend toward "outcome-based" advertising, where platforms optimize the user experience and performance using AI, even if it sidelines advertiser preferences. Third, the timing — just ahead of the 2026 holiday season — is strategic. E-commerce brands are already allocating budgets, and a new, visually engaging ad surface in a high-attention environment like ChatGPT could draw meaningful spend away from established players like Google and Meta.

From a competitive standpoint, Google Shopping should watch ChatGPT’s moves closely. While ChatGPT’s ad business is still nascent, the product carousel, coupled with its massive user base, positions the AI assistant as a direct competitor in product discovery. If OpenAI opens the carousel to multi-retailer placements, it becomes a head-to-head battleground for the very same retail search dollars that currently flow to Google. This is especially potent given that ChatGPT’s conversational context often implies higher purchase intent — users are explicitly asking for advice, comparisons, or recommendations. The AI assistant’s ability to natively blend answer and advertisement could make its product carousel an exceptionally effective conversion tool.