Samsung Electronics deploys ChatGPT and Codex across 100K+ workers globally

Samsung’s largest AI rollout signals mainstream enterprise adoption

Samsung Electronics is deploying ChatGPT Enterprise and Codex to its entire Korean workforce and all employees in its Device eXperience (DX) division globally, according to the OpenAI Blog. According to OpenAI, this represents one of the company’s largest enterprise deployments on record, covering technical and non-technical roles across research, development, manufacturing, product design, marketing, and corporate operations.

The scale of Samsung’s adoption reflects a broader shift: AI-assisted productivity tools are no longer confined to engineering teams or proof-of-concept pilots. Instead, Samsung is treating ChatGPT and Codex as foundational platforms for knowledge work across the organization, with employees using the tools for document drafting, data analysis, code generation, and workflow automation.

Codex expands beyond code into enterprise workflows

Codex, which originated as a code-generation assistant, is now serving dual duty at Samsung. According to the OpenAI Blog, the tool helps developers write, review, and debug software while enabling non-technical teams to automate internal processes and build custom applications without writing code themselves. The expansion mirrors a market-wide pattern: code-as-interface has become a bridge between business logic and implementation.

Samsung’s deployment comes as Codex use has accelerated in Korea specifically. According to OpenAI, weekly active Codex users in Korea grew nearly 800% between February 1, 2026 and the time of Samsung’s announcement, suggesting strong demand for AI-assisted development tools in the region. More broadly, Codex now has over 5 million weekly active users across technical and non-technical roles.

Enterprise security and governance layers unlock adoption

ChatGPT Enterprise provides the data residency, access controls, and governance frameworks that large organizations require to deploy consumer-grade AI tools at scale. According to the OpenAI Blog, Samsung is using these enterprise-grade protections to ensure employee usage complies with company security policies. This abstraction—where OpenAI handles the baseline security layer—lowers the barrier for IT and legal teams to approve broad deployment.

The availability of these controls suggests that enterprise hesitation around generative AI has shifted from feasibility to operational scaling. Samsung’s decision to deploy across Korea plus a global division is not a pilot; it is a production commitment.

Samsung-OpenAI infrastructure partnership deepens

The software deployment is paired with hardware collaboration. According to OpenAI, Samsung Electronics is supplying advanced memory semiconductors to OpenAI for next-generation AI infrastructure, expanding the relationship beyond licensing and into the supply chain for large-language-model training and inference. This two-way partnership—Samsung consuming OpenAI software while supplying chip technology—reflects the consolidation of AI value chains around leading model providers and semiconductor makers.

Why This Matters

Samsung’s scope of deployment—entire workforce in one region, global division in another—resets expectations for enterprise AI adoption timelines. If Samsung has moved from evaluation to production across 100,000+ employees within a year of ChatGPT’s release, other conglomerates face pressure to accelerate their own rollouts or risk internal productivity gaps.

For OpenAI, the deployment validates demand for ChatGPT Enterprise as an enterprise-grade SaaS offering rather than a consumer upsell, and it demonstrates Codex’s versatility beyond pure software development. The 800% quarterly growth in Korean Codex users suggests regional regulatory or cultural conditions may favor rapid adoption—a signal for other software companies expanding into Asia.

For Samsung itself, the move signals confidence that large-language models are now reliable enough for mission-critical workflows while remaining flexible enough to serve both engineering and business functions. The next question is adoption velocity: whether 100,000 employees will shift their workflows materially within six months, or whether broad access will mirror broad skepticism.