ChatGPT's shopping updates are a play for the messy middle of product discovery

OpenAI’s push into commerce is ostensibly about shopping, but the critical operating word is discovery. Checkout is merely the endpoint; discovery is the swamp where users compare options, weigh constraints, and inadvertently become experts in dishwasher dimensions.

The ChatGPT product discovery announcement details richer visual browsing, side-by-side comparisons, image-based inspiration, and conversational refinement. By expanding the Agentic Commerce Protocol, OpenAI is allowing merchants to inject up-to-date catalog data directly into the chat interface.

This represents a much larger ambition than simply putting a 'buy' button in a chat window. A conversation is genuinely well-suited to fuzzy shopping queries where a user knows their budget and constraints but wants to avoid opening a dozen browser tabs. ChatGPT is trying to own the decision layer.

The risk, however, requires a neon warning sign: discovery can quickly devolve into paid-placement sludge if the incentives blur. Users might tolerate sponsored recommendations if they are clearly disclosed, but they will violently reject a helpful shopping assistant that quietly operates as an affiliate maze with excellent manners.

For this to become a lasting habit, the AI must clearly explain why products match the user’s needs and allow for fluid adjustments without restarting the search. If it effectively reduces the tab swamp, OpenAI wins influence over the transaction without ever having to process the credit card.