Samsung opens ChatGPT and Codex to Korea staff as Codex usage jumps 800%

Samsung’s AI rollout is moving from team-level experimentation to company-wide operating infrastructure. Under a new deployment, ChatGPT Enterprise and Codex will be available to all Samsung Electronics employees in Korea and to Device eXperience (DX) employees worldwide.

The scale is the news. OpenAI describes the agreement as one of its largest enterprise deployments to date. It also says more than 5 million people now use Codex every week across technical and non-technical work, while Codex weekly active users in Korea have grown nearly 800% since February 1, 2026. Those numbers suggest coding agents are becoming a broader workplace layer, not just a developer-side productivity tool.

Samsung plans to use ChatGPT and Codex across software development, product development, manufacturing, marketing, and corporate functions. ChatGPT Enterprise gives the company data protection, access management, and security controls. Codex is framed not only as a coding assistant for writing, reviewing, and debugging software, but also as a way for non-technical teams to turn ideas into internal tools, websites, and automated workflows.

The deployment also deepens a relationship that already touched AI infrastructure. Samsung Electronics has been working with OpenAI on supplying advanced memory semiconductors for future AI systems. That makes Samsung both a supplier to the AI infrastructure buildout and now one of the largest enterprise customers trying to make AI tools part of daily operations.

The practical question is governance. A deployment across engineering, manufacturing, and business teams creates more value only if employees can use the tools inside clear data, audit, and review boundaries. For a company that handles product roadmaps, manufacturing processes, and sensitive commercial data, the interesting test is not whether workers open ChatGPT. It is whether AI-assisted work can be measured, reviewed, and kept inside Samsung’s security policies while still speeding up execution.