Samsung SDS, the IT services arm of the South Korean conglomerate, will begin reselling OpenAI’s ChatGPT, Google’s Gemini, and Anthropic’s Claude to Samsung Group affiliates by the end of June 2026, embedding strict governance controls for enterprise-wide adoption. The move positions SDS as the internal broker and operator for a sweeping AI transformation across the group’s diverse business units—spanning electronics, biopharma, construction, and financial services—while keeping compliance, security, and data sovereignty firmly in corporate hands.

The announcement, disclosed in internal planning documents seen by Windows News, outlines a phased rollout that turns SDS into a centralized AI SaaS hub. Rather than each affiliate separately negotiating and integrating with disparate AI vendors, SDS will procure, package, and manage the models under a unified governance framework. This is a pivot from Samsung’s past caution: in 2023, the company briefly banned employee use of ChatGPT after an internal data leak, only to subsequently develop its own enterprise AI offerings. Now, by acting as gatekeeper, SDS aims to harness public frontier models securely, ensuring that proprietary data never leaves Samsung’s designated environment.

Governance is the linchpin. SDS will enforce role-based access controls, data loss prevention, audit trails, and prompt filtering, aligning with South Korea’s tight data regulations and Samsung’s internal security posture. The service will be integrated with Samsung’s existing single sign-on and identity management systems, many of which are deeply tied to Microsoft Azure Active Directory and Windows Hello for Business authentication. For the thousands of Windows-based endpoints across Samsung’s offices, this means IT administrators can push AI app configurations, set usage policies, and monitor interactions via group policy objects and Microsoft Intune—a direct plug into the Windows administration ecosystem.

Multi-Model Strategy: Why Three Models?

SDS is not betting on a single AI provider. Instead, it is orchestrating a trio of the most powerful large language models, each serving different use cases:
- OpenAI ChatGPT (likely the latest GPT-4 Omni or future successor) will power conversational interfaces, customer support bots, and content generation tasks across marketing and HR.
- Google Gemini is expected to shine in data analysis, research-heavy workflows, and integration with Google Workspace, which many Samsung affiliates use alongside Microsoft 365.
- Anthropic Claude, with its constitutional AI safety guardrails, will be positioned for sensitive legal, compliance, and ethical review tasks within Samsung’s heavily regulated financial and life sciences divisions.

By offering all three, SDS gives business units a “model cafeteria” where cost, capability, and compliance can be balanced. A product design team might use Gemini’s multimodal reasoning to analyze technical schematics, while a legal team simultaneously uses Claude to review contracts—all under SDS’s logging and governance umbrella. This approach also mitigates vendor lock-in and provides redundancy if one model evolves in an unfavorable direction or increases pricing.

The Role of Windows Administration

For Windows admins, this isn’t just another SaaS app. Samsung SDS plans to deliver these AI tools through a managed desktop experience on Windows 11 Enterprise, leveraging the Microsoft Endpoint Manager suite. Admins will define AppLocker policies to whitelist only the SDS-provisioned AI clients, preventing shadow IT. Conditional Access policies in Azure AD will ensure that AI chats can only happen from compliant, corporate-managed devices. Data inputs and outputs will be scanned with Microsoft Purview Information Protection, labeling and encrypting sensitive content before it ever reaches the model provider.

SDS is also building a dashboard—likely hosted on Azure and connected via Windows Admin Center—that shows real-time usage analytics, cost allocation, and incidents across all affiliates. This gives CIOs a macro view of AI consumption while allowing local IT teams to drill down to individual user prompts for compliance reviews. Early pilot users within Samsung Electronics are already testing the integration on Windows 11 24H2, with the provisioned apps appearing in the Company Portal alongside traditional software.

Timeline and Rollout

According to the planning documents, the project—codenamed “Bridge” internally—has three phases:
- Phase 1 (Q1 2025 – Q2 2025): Infrastructure setup, vendor contracting, and security architecture review. SDS is currently negotiating enterprise agreements with OpenAI, Google Cloud, and Anthropic to define data processing terms and geo-residency (all data is expected to stay within South Korea’s borders).
- Phase 2 (Q3 2025 – Q1 2026): Pilot programs in select affiliates: Samsung Electronics (mobile division), Samsung Biologics, and Samsung Fire & Marine Insurance. Windows management policies will be stress-tested during this stage.
- Phase 3 (Q2 2026): General availability for all Samsung Group affiliates. By June 30, 2026, any business unit should be able to onboard users through a standard service request.

This deliberate pace underscores the complexity of marrying consumer-grade AI with ironclad enterprise controls. Samsung’s sprawling conglomerate structure—with over 20 major affiliates—means that SDS must cater to everything from chip fabrication clean rooms where security is extreme, to retail showrooms where employee turnover is high.

Security and Sovereignty: The Samsung Way

Samsung’s approach mirrors a broader industry trend: enterprises crave generative AI’s productivity gains but fear data leakage and regulatory penalties. In Korea, the Personal Information Protection Act (PIPA) imposes hefty fines for mishandling personal data, and cross-border data transfers are strictly regulated. SDS’s governance layer will ensure that prompts and responses are stripped of personally identifiable information (PII) before they leave the local network, and that all queries are passed through a self-hosted proxy that can redact, block, or log based on keyword dictionaries.

Moreover, Samsung is likely to use its own Knox security platform—originally built for mobile devices—as an additional seal on Windows endpoints. The Knox framework can enforce device-level encryption and integrity checking, meaning that even if a user tries to paste sensitive code into ChatGPT, the system can intercept and warn. For affiliates still using Windows 10, SDS will mandate an upgrade to Windows 11 by early 2026 to provide the needed virtualization-based security and hardware-enforced stack protection.

Competitive Landscape and Market Impact

SDS isn’t alone in this space. Global system integrators like Accenture and Deloitte already offer “AI broker” services, but Samsung’s move is distinct because it first targets internal scale before potentially commercializing the platform to external Korean enterprises. If successful, SDS could emerge as a dominant enterprise AI governance provider in East Asia, leveraging its deep integration with Samsung’s sprawling manufacturing and services ecosystem.

For Microsoft, the news is a double-edged sword. On one hand, SDS’s heavy reliance on Azure AD, Intune, and Windows 11 strengthens the Microsoft ecosystem within Samsung. On the other, by offering Claude and Gemini, SDS validates a multi-model world where enterprises are not content to rely solely on Microsoft Copilot. In fact, Samsung is a major Microsoft 365 customer, but the ability to pick the best model for each task—without sacrificing governance—may spur other conglomerates to demand similar flexibility from their IT providers.

What Comes Next

Looking beyond 2026, SDS envisions a “model orchestration” layer that can dynamically route prompts to the most suitable engine based on cost, latency, and accuracy. It is already experimenting with fine-tuning Gemini on proprietary manufacturing data for predictive maintenance, while using Claude to enrich customer call transcripts with sentiment analysis. The ultimate goal is a self-service AI marketplace where any approved employee can spin up a custom copilot as easily as requesting a virtual machine—always under the watchful eye of the SDS governance console.

For Windows administrators, the message is clear: enterprise AI adoption is no longer a question of “if” but “how tightly managed.” SDS’s blueprint—centralized procurement, unified policy enforcement, and deep Windows integration—could become the template for every large organization that wants to empower employees with AI while keeping corporate secrets safe. The countdown to June 2026 has begun, and Samsung’s internal IT ecosystem is the testbed for the governed AI era.