OpenAI taps Japan’s Datasection to expand across Asia

Datasection on May 29 revealed that its TAIZA platform has been integrated with OpenAI’s API to manage enterprise workloads using a controlled deployment model. This setup will allow the system to route requests to OpenAI’s models via Datasection’s software layer. It won’t have to directly access OpenAI’s cloud endpoints.

This comes in as the Japanese neocloud operator ramps up GPU capabilities in several nations. This integration may indicate OpenAI’s first deployment of model inference on a non-Microsoft cloud setup.

OpenAI pushes beyond Microsoft cloud

OpenAI and Microsoft had an updated partnership agreement signed in April 2026. It allowed the AI giant to distribute its products across various cloud providers. However, it’ll have to maintain Microsoft as its primary cloud partner. It grants the software company non-exclusive licensing rights to OpenAI’s models until 2032.

The updated partnership has removed the limitations imposed. Earlier, it confined OpenAI largely to Microsoft Azure. This enabled a broader distribution among other cloud providers like AWS and Google Cloud, where needed. As of now, the AI company can serve all of its products to customers across any cloud provider.

Microsoft will still continue to host a huge amount of OpenAI’s workload on Azure and remains a major shareholder. The Datasection partnership appears to align with the API-level distribution model rather than any alteration in OpenAI’s infrastructure deployment. Its primary method for delivering its model will be through the application programming interface.

OpenAI has also been expanding its regional data residency options across Asia since 2025. This includes Japan, Singapore, India, and South Korea. Its integration with Datasection is seen as a trend of models being embedded into third-party enterprise platforms.

Datasection shares jump 20% amid OpenAI deal

The Japanese company is reportedly increasing its AI computing capacity in Australia and Thailand, too. It is building what it calls a multi-country GPU infrastructure network. As of May 2026, its GPU capacity exceeded 20,000 units.

Its AI infrastructure business started in September 2025. The rapid growth in monthly sales from October came in and hit 4.5 billion Japanese Yen (approx $28.25 million). The company’s expansion completely relies on NVIDIA-based systems. It holds 5,000 GPUs in Japan, 10,000 in Australia, and 5,000 in Thailand.

Data from the Tokyo Stock Exchange shows that the Datasection share price jumped by almost 20% in a day. It traded at 6,140 Yen at the press time. The company’s AI strategy is to consolidate computing infrastructure and enterprise deployment tools into a single platform.

Datasection CEO Norihiko Ishihara stated that the partnership with OpenAI marks a crucial step in aligning regional infrastructure with evolving sovereign AI requirements. The collaboration aims to connect advanced AI systems with the needs of enterprise and public-sector deployments across the Asia-Pacific.

OpenAI has not released any separate statement regarding the Datasection integration. In its public disclosures, the company continues to stress API-based distribution and partner-led deployment as its core enterprise strategy.