Samsung Electronics, which banned ChatGPT internally in 2023, is now deploying it across its workforce in what OpenAI calls one of its largest enterprise rollouts to date. OpenAI announced on June 21 that ChatGPT Enterprise and its Codex coding tool will go to every Samsung Electronics employee in South Korea, and to all employees worldwide in the company's Device eXperience division, the unit behind its phones, TVs, and home appliances.
The reversal is the heart of the story. Samsung's 2023 ban came after an employee pasted internal source code into ChatGPT, raising fears that sensitive data could leak into a public model. What makes the about-face possible now is ChatGPT Enterprise, whose data-protection, user and access management, and security controls let Samsung run the tools inside its own security and governance rules rather than on the open consumer service. The bridge from ban to company-wide deployment was built out of enterprise controls.
On the ground, the two tools split the work. Employees will use ChatGPT for knowledge tasks, searching and analyzing information, drafting documents, developing ideas, and interpreting data, and Codex for software, writing, reviewing, and debugging code and turning ideas into internal tools, websites, and automated workflows. Samsung says it plans to use them across R&D, manufacturing, marketing, and corporate functions, which for a company that ships this much hardware and software is a wide surface area.
The scale is what makes it notable. Samsung Electronics is one of the largest technology employers in the world, so handing ChatGPT and Codex to all of its Korean staff and its global device division at once is a flagship enterprise deal for OpenAI, and a marker of how far corporate adoption of frontier AI has moved. Enterprise AI has spent two years stuck between cautious pilots and outright bans, and a rollout this broad is a sign that the default posture at large companies is shifting toward giving the tools to everyone.
The honest caveat is that a deployment is an intent and an access grant, not a measured result. Whether Samsung actually gets the productivity it is buying will take time to show, and the 2023 leak that triggered the original ban is a reminder that enterprise controls reduce the risk of mishandled data without erasing it. But the symbolism is hard to miss. The same company that forbade ChatGPT two years ago over a single leak is now putting it in front of its entire workforce, which says as much about the speed of the shift in enterprise AI as any benchmark could.