Samsung already makes the chips that power AI. Now, it reportedly wants to give some of the industry’s biggest AI developers a place to put them.
The group’s IT services arm Samsung SDS is in preliminary discussions with OpenAI and Anthropic about becoming tenants at its AI data centers, according to the Financial Times. No agreements have been announced, but the talks come as Samsung expands its AI infrastructure investments and develops a major data center in South Korea.
For channel partners, the expansion could create opportunities in infrastructure deployment, integration, and ongoing support. It also raises questions about how much of that business Samsung will handle through its own subsidiaries and services operations.
Samsung pursues AI data center tenants beyond its chip business
The discussions are part of Samsung’s plan to bring together capabilities across its subsidiaries, including semiconductor manufacturing, construction, power infrastructure, and data center operations.
Samsung has already committed significant funding to this strategy.
In September, Samsung Electronics and five affiliates announced a combined $1 billion investment in Helix Digital Infrastructure, a KKR-backed company developing AI data centers, power systems, and connectivity infrastructure.
Samsung’s investment follows KKR’s June launch of Helix Digital Infrastructure, which focuses on developing AI-ready data centers and supporting infrastructure.
For infrastructure partners, projects of this scale may create demand for specialized equipment, deployment services, and ongoing maintenance. However, Samsung has not announced which portions of its planned data center work will be available to external channel companies.
Gumi AI data center prepares for high-density GPU deployments
Samsung SDS is already constructing an AI data center in Gumi, South Korea.
According to its announcement, the facility will initially support 60 megawatts of power capacity, with plans to expand to as much as 120 MW.
Construction is scheduled for completion by the end of 2028.
The facility will use a combination of air and liquid cooling to support high-performance GPU servers. Samsung SDS plans to operate its own GPU infrastructure while also offering colocation services that allow customers to install their own equipment.
Samsung has not confirmed whether Gumi is among the facilities being discussed with OpenAI or Anthropic.
For partners working with enterprise customers, the project offers an example of the infrastructure requirements behind larger AI deployments.
Rising AI data center power demand could create opportunities for MSPs to help customers improve infrastructure efficiency and plan capacity upgrades.
Those services may become increasingly important as AI customers evaluate power availability, cooling requirements, and operating costs alongside GPU performance.
Samsung’s existing AI partnerships could increase channel competition
The reported tenancy discussions would build on relationships Samsung already has with both AI developers.
In October 2025, Samsung and OpenAI announced a strategic partnership involving AI infrastructure and enterprise services. Samsung SDS was also authorized to resell OpenAI’s enterprise offerings in South Korea.
Samsung has also expanded its partnership with OpenAI to include next-generation AI chip development.
Samsung SDS also signed a strategic partnership with Anthropic in July 2026, covering enterprise Claude deployments, joint business development, and training for engineers who implement AI solutions.
These relationships give Samsung SDS a presence in both infrastructure and enterprise AI implementation.
For independent MSPs and systems integrators, that creates a potential challenge. Samsung could supply the infrastructure and compete for the consulting, deployment, and support contracts surrounding it.
Partners will need to determine where they can offer services that Samsung does not already provide directly.
Channel partners should assess capacity, contracts, and service opportunities
Samsung’s expansion could benefit channel companies that support enterprise AI infrastructure, but there are no confirmed resale opportunities tied to the reported discussions.
Partners evaluating future projects should focus on several questions:
- Infrastructure services: Will Samsung use external partners for networking, cooling, power management, or equipment maintenance?
- Resale and integration rights: Can MSPs resell infrastructure capacity or enterprise AI services, and what margins and support responsibilities would apply?
- Customer requirements: Can partners support clients with hybrid deployments, security controls, data residency, and integration with existing IT systems?
- Delivery commitments: What capacity is confirmed, when will it become available, and which service-level agreements will apply?
For MSPs, a practical near-term opportunity is helping customers assess whether their networks, storage, security controls, and budgets can support AI workloads.
In the near term, MSPs and systems integrators should focus on helping enterprise customers assess whether their networks, storage, security controls, and budgets can support AI workloads—and prepare to connect hosted AI tools with on-premises systems.
However, partners should avoid promising Samsung-backed capacity or building offerings around the reported OpenAI and Anthropic talks until commercial terms are confirmed.
If Samsung finalizes agreements with major AI developers, the resulting buildouts could create significant downstream services work. How much independent channel companies benefit will depend on confirmed commercial terms, available partner roles, and how much of the deployment Samsung keeps in-house.
As Samsung expands its AI infrastructure business, channel partners need to be ready to help customers put AI to work. Build your skills with The Neuron Academy’s free six-minute How to Talk to AI course, explore more AI lessons, and get seven days of free access.


