Why OpenAI’s Altman is headed to Samsung, and what’s at stake

Why OpenAI’s Altman is headed to Samsung, and what’s at stake
Devesh Kumar
11 Jun 2026, 12:13 PM

powered by

Invezz
Samsung Electronics (005930.KS)

Buy. Altman’s Suwon visit is execution, not hype: it signals Samsung will deepen supply and co-build AI infrastructure (memory + data-center + enterprise AI). That supports sustained high-end memory demand and strengthens Samsung’s “full-stack” positioning versus peers. Expect multiple expansion if Samsung converts LOIs into visible, recurring AI-related revenue.

Key Risk: OpenAI’s Stargate timelines or spend get delayed/cut, leaving Samsung with memory demand that normalizes faster than expected.

SK Hynix (000660.KS)

Buy. The article ties SK Hynix to advanced memory supply for OpenAI’s data centers. If Samsung is the platform, SK Hynix is the bottleneck supplier for AI memory—so any follow-through on Stargate increases probability of tighter supply and better pricing.

Key Risk: AI memory demand weakens (or competitors add capacity) and SK Hynix pricing power fades despite the partnership talk.

  • Sam Altman to visit Samsung’s Suwon campus on June 15.
  • OpenAI and Samsung deepen AI chip and data-centre ties.
  • Samsung pushes AI-driven factories and workplace automation.

OpenAI CEO Sam Altman is set to visit Samsung Electronics’ Suwon campus on Monday, in a trip that signals a deeper push by the ChatGPT maker into Asia’s corporate and hardware ecosystem.

The visit is expected to bring together one of the world’s most influential AI companies and one of its most important chip and device makers.

For OpenAI, Samsung offers memory chips, data-centre expertise and global hardware reach.

For Samsung, Altman’s visit comes as the company tries to put AI at the centre of its workplace, factories and future growth story.

From letters of intent to real action

The trip follows Altman’s high-profile visit to Seoul in October 2025, when OpenAI signed a series of agreements with Samsung and SK Group tied to its Stargate AI infrastructure project.

Those agreements covered advanced memory chips, data-centre development, enterprise AI services and even potential floating data centres.

Samsung Electronics and SK Hynix had signed letters of intent to supply memory chips for OpenAI’s data centres, while Samsung SDS, Samsung C&T and Samsung Heavy Industries were also drawn into the partnership.

OpenAI said the broader plan was to expand global AI infrastructure and increase the supply of advanced memory chips needed for next-generation AI models.

The October agreements created the framework and this visit looks more like a follow-up on execution.

According to Seoul Economic Daily, Altman is expected to discuss cooperation across mobile AI, device-based AI services, semiconductors and data centres.

Samsung’s AI overhaul, and why it needs OpenAI

Samsung’s interest in OpenAI is not limited to chips. The company is trying to push AI deeper into its own operations, from office work to manufacturing.

The Asia Business Daily reported that Altman is scheduled to speak at Samsung’s “DX Insight Talk” event at the Suwon campus, with a lecture focused on AI-based work innovation.

That timing is important as Samsung has already said it wants to transform its global manufacturing operations into “AI-driven factories” by 2030.

The plan includes digital twins, specialised AI agents and more automation across production, quality control and logistics.

Altman’s lecture gives Samsung’s employees and executives a direct view of how OpenAI thinks about AI inside large organisations.

The stakes are also financial as Samsung posted record first-quarter results this year, helped by strong demand for memory chips used in AI infrastructure.

But it still faces pressure to prove that it can stay central to the next phase of the AI boom, not just as a supplier of components, but as a full-stack technology partner.

That is where OpenAI becomes useful. It gives Samsung a high-profile customer, a potential enterprise AI partner and a window into where demand for future computing power may be heading.