OpenAI and Samsung partner on AI chip development

OpenAI has deepened its partnership with Samsung Electronics to include joint research on next-generation AI chips alongside expanding enterprise AI services, according to details first reported by Reuters.

Harrison Kim, General Manager of OpenAI Korea, disclosed the chip collaboration at a Seoul press conference Wednesday. "One of the areas where we have made the most progress and gained the most recognition with Samsung Electronics is our joint production and research on the next-generation chips we are developing," Kim said, though he declined to provide specifics. Samsung Electronics did not comment on the partnership.

Enterprise AI deployment scales rapidly

The partnership extends beyond semiconductors into enterprise software adoption. Kim described Samsung Electronics as "one of the largest-scale deployments of ChatGPT globally," with employees across research and development, marketing, and sales using the platform in South Korea and internationally.

ChatGPT Enterprise usage across South Korean businesses and institutions grew approximately 28-fold by late August compared to the same period the previous year, according to OpenAI. The company did not disclose absolute user numbers but noted that as of late last year, no country outside the United States had more paying ChatGPT subscribers than South Korea.

Memory chip demand and supply agreements

Kim emphasized that advancing AI hardware sophistication will continue driving demand for memory chips, making close collaboration with South Korean semiconductor suppliers a core focus for OpenAI's local office. Last year, both Samsung Electronics and SK Hynix signed letters of intent to supply memory chips for OpenAI's Stargate data center project.

Samsung has begun shipping samples of its HBM4E high-bandwidth memory chip to major customers, becoming the first company to distribute this next-generation AI memory product. These chips are used in AI accelerators including Nvidia's Rubin and Google's Ironwood Tensor Processing Unit. Samsung's customer base for these components includes AMD, Nvidia, and Google.

OpenAI's custom chip strategy

OpenAI unveiled its first custom AI chip, Jalapeno, in June. Built in collaboration with Broadcom to handle AI inference—the process of generating responses in tools like ChatGPT—the chip will be manufactured at TSMC facilities in Taiwan.

In June, Samsung announced that employees within its Device eXperience division, which handles mobile phones, televisions, and home appliances, received approval to use ChatGPT, Google's Gemini, and Anthropic's Claude in daily work. OpenAI expects AI adoption to accelerate following the September 4 launch of GPT-6 Astra, its latest model.

Why it matters

The partnership signals OpenAI's strategy to secure semiconductor supply chains while building custom hardware, reducing dependence on third-party chip providers. For Samsung, the collaboration offers a pathway into AI infrastructure beyond memory chips, potentially positioning the company as a more integrated AI hardware supplier as enterprise adoption accelerates across Asia.

Details of the partnership were first reported by Reuters.