"In the offline era, consumers picked one product from dozens on a shelf. In the digital era, they had to directly compare thousands or tens of thousands of products, and algorithms emerged to help that choice. In the coming era of artificial intelligence (AI) shopping agents, the options may again be compressed drastically, and further, the structure could change so that the agent even completes the purchase. Corporations that respond swiftly to the new change will seize the opportunity."
On the 16th at the Westin Josun Hotel in Jung-gu, Seoul, James Jang, CEO of Gmarket Co., said this in a keynote speech titled "Conditions for choice in the AI era: Changes in consumer standards and market order reshaped by AI" at the 2026 ChosunBiz Consumer and Retail Forum. Jang said, "If AI comes to handle not only search, comparison, and recommendation in shopping but also the purchasing process, the very way of consolidation between consumers and products could change."
Jang focused on the "AI shopping agent" among the various impacts AI will have on e-commerce. He said AI agents specialized for shopping will take on the role of finding, comparing, and recommending products on behalf of customers.
He said, "Existing algorithms worked to aid product exposure based on accumulated search terms, clicks, and purchase histories within the platform," adding, "AI agents can synthesize various signals such as product information and reviews, price, seller reliability, and customer preferences to present fewer options and even handle the purchase."
In this case, he said, the way brands and manufacturers are chosen by consumers will also change. Jang said, "AI agents do not respond to emotional elements," adding, "People can be influenced by product pages, advertisements, reviews, and influencer recommendations, but AI judges based on a product's objective advantages, the consistency of reviews, and the fit with consumer preferences."
Jang stressed that brands and manufacturers should rebuild product information in ways AI can read. He said, "A product's differentiated functions and value proposition must be clear, and details, metadata, and reviews must be structured so AI can easily recognize them," adding, "If search engine optimization (SEO) was important before, AI optimization will become even more necessary going forward."
However, Jang predicted AI will not replace every shopping experience. While areas with many repeat purchases, such as daily necessities or grocery shopping, may be quickly entrusted to AI, in areas where taste and enjoyment matter—like fashion, cosmetics, and food ingredients—the experience of people picking directly may remain.
"Shopping is fundamentally enjoyable," Jang said. "Consumers can leave products they are less interested in to AI and dig deeper into the fields they love. From a platform's perspective, strategies must be set with users' preferences in mind."
Jang also pointed out hurdles that must be overcome as AI-based shopping spreads. He said that although AI shopping agents are already close in technical terms, there is a psychological barrier for consumers to immediately entrust purchasing authority and their wallets.
He said, "Most AI today remains at the stage of a research helper that gathers and compares information," adding, "More trust is needed for consumer confidence to build. Just as people assign work to a new hire, as trust in AI shopping agents accumulates, consumers will entrust them with more important purchases."
He also addressed AI's inherent limits. Because AI cannot directly use or experience products, its recommendation accuracy may be limited for new products with little data or for products where user experience is crucial. He also cited as a problem that when incorrect information—such as fake reviews, false advertising, or fabricated transaction volumes—is fed in, it is hard to filter it out completely.
Despite such limits, Jang viewed Korea as the market where AI commerce changes could emerge the fastest. He explained, "In South Korea, not only shopping but also transportation, services, travel, and administration have digitized rapidly, and data spanning all age groups has been accumulated," adding, "It is an excellent environment for AI to be introduced for the first time."
Jang said, "Whenever a new technology emerges and the paradigm shifts, new winners appear beyond the companies that had been doing well," adding, "Ultimately, more important than which corporations or which industries is how quickly you respond to change. In the AI era, anyone can be a winner."