OpenAI Korea’s First Year: Enterprise Users Up 28-Fold

At a September 9, 2026, briefing marking its local office’s first anniversary, OpenAI Korea described a market moving from individual ChatGPT use toward wider deployment inside organizations. Its headline figures were roughly 28-fold year-over-year growth in Korean ChatGPT Enterprise users as of the end of August and 35 percent growth in the country’s overall ChatGPT user base. 71213

What the growth figures measure

The two figures describe different groups. The 28-fold increase refers to people using ChatGPT Enterprise at Korean companies and institutions—not a 28-fold increase in the number of corporate customers or in revenue. The 35 percent figure covers the broader Korean ChatGPT user base. They should not be read as interchangeable measures of business adoption. 267

OpenAI Korea’s head, Kim Kyung-hoon, said some companies had bought ChatGPT Enterprise through OpenAI’s headquarters before the local office opened, but adoption by companies increased after it established a Korean presence. That account gives the office a role in supporting growth; the reported figures alone do not show how much of the increase it caused. 4

From chat assistance to organizational work

The briefing’s clearest explanation for the enterprise surge was a change in how businesses used AI. Reporting described a move beyond information searches and document drafting toward coding, research, analysis and use across company operations. Korean weekly active users of ChatGPT Work and Codex also rose about 14-fold from March to August 2026, a separate measure over a shorter period. 2513

OpenAI cited Samsung Electronics and Seoul National University as examples of large-scale adoption. Reporting says Samsung made ChatGPT Enterprise and Codex available to employees in Korea and overseas. The supplied accounts do not establish the precise reach of that deployment or verify a figure of nearly 50,000 university users, so neither should be treated as a confirmed measure of adoption. 127

The second-year infrastructure question

Deeper corporate use also makes the location of data and computing more consequential. OpenAI already offered data residency—storage in Korea—for ChatGPT Enterprise, ChatGPT Edu and API customers. Its proposed inference residency would go further by carrying out in Korea the GPU computation used to process requests and generate responses. Kim said OpenAI was discussing domestic inference infrastructure with potential partners; the reporting does not establish that local inference was already running or that a data-center plan was finalized. 81415

Kim also said OpenAI would prioritize partnerships with Korean companies in the semiconductor supply chain, while reporting pointed to cooperation with Samsung on next-generation chips. Those discussions are distinct from a confirmed domestic inference deployment. 111 The first-year result, then, is measurable growth in users and workplace use; the next test is whether OpenAI can turn its local processing plans into an available service. 7814