# Samsung Puts ChatGPT and Codex in Front of Its Global Workforce

One of OpenAI's largest enterprise deployments is going to a company that banned generative AI internally three years ago.

- Published: 2026-06-22T10:48:41.833Z
- Canonical: https://polylog.news/ai/2026-06-22/samsung-puts-chatgpt-and-codex-in-front-of-its-global-workfo
- Publisher: Polylog (AI desk)
- Section: tech
- Sources: [OpenAI](https://openai.com/index/samsung-electronics-chatgpt-codex-deployment), [The Korea Times](https://www.koreatimes.co.kr/business/tech-science/20260622/openai-lands-samsung-as-major-chatgpt-enterprise-customer)

Samsung Electronics is deploying ChatGPT Enterprise and Codex across its workforce, a rollout OpenAI [describes as one of its largest enterprise deployments to date](https://openai.com/index/samsung-electronics-chatgpt-codex-deployment). Access goes to all employees in Korea and to the full global headcount of its Device eXperience division, spanning software development, marketing, product work, and manufacturing.

The Codex component is the part engineers should note. Beyond writing, reviewing, and debugging code, Samsung is positioning it for non-technical staff to assemble internal tools, websites, and automated workflows. This is the agentic-coding pattern moving from developer pilots into general enterprise use.

The history matters here. Samsung banned generative-AI tools internally in 2023 after engineers leaked source code and meeting notes through ChatGPT. The reversal rests on ChatGPT Enterprise's data-isolation, access-management, and security controls, the features that make the difference between a pilot and a governed deployment. The [Korea Times framed the agreement](https://www.koreatimes.co.kr/business/tech-science/20260622/openai-lands-samsung-as-major-chatgpt-enterprise-customer) as a major enterprise win for OpenAI in a competitive market.

The unverified part is impact. A seat count is not a productivity number, and Samsung has not published measured gains. The signal is on the procurement side. A company previously harmed by a data leak judged the governance controls sufficient to standardize on a frontier vendor.

## What this means

Enterprise adoption at this scale is decided by data-governance and access controls more than by benchmark scores, and a company that previously banned these tools clearing them is a stronger adoption signal than another startup case study. The open question is whether agentic coding for non-engineers produces durable output or shifts the maintenance burden onto the teams who inherit auto-generated tools.

## What to watch

- Whether Samsung or OpenAI later publishes measured productivity or defect-rate data, which would move this from procurement signal to evidence.
- How much of the Codex usage comes from non-developers, the real test of agentic coding's reach beyond engineering.