Microsoft Corporation (MSFT.US) reaffirms relationship with OpenAI: pivoting towards AI self-sufficiency and focusing on developing cutting-edge models in-house.

Mustafa Suleiman, head of the artificial intelligence business at Microsoft Corporation (MSFT.US), said that the company is seeking to achieve "true self-sufficiency" in the field of AI by independently developing high-performance models and gradually reducing its dependence on OpenAI's technology. Suleiman revealed that this strategic shift began in October 2025 after the restructuring of the partnership between Microsoft Corporation and OpenAI. Since then, Microsoft Corporation has started independently building cutting-edge technology, rather than relying on external partners. Microsoft Corporation has invested over $13 billion in OpenAI to date and holds nearly 27% ownership in the ChatGPT developer. "We must develop our own foundational models - standing at the absolute forefront, possessing exascale computing power, and having a world-class AI training team," Suleiman said. He is one of the co-founders of Alphabet Inc. Class C DeepMind and joined Microsoft Corporation in March 2024. Suleiman stated that Microsoft Corporation is heavily investing in the integration and organization of large-scale datasets, which are essential for training advanced systems. "This is the core mission for us to truly achieve self-sufficiency." Previously, Microsoft Corporation relied on OpenAI's models to power its own AI services like Copilot. According to a new agreement reached in October 2025, Microsoft Corporation's intellectual property rights at both the product and model ends have been extended until 2032, expanding to include related technologies after general artificial intelligence, with corresponding security guards. At the same time, OpenAI also gained more freedom to introduce new investors and infrastructure partners. In addition to OpenAI, Microsoft Corporation has also invested in AI companies like Anthropic and Mistral from France. However, internally developed models are progressing rapidly. Suleiman stated that self-developed products are expected to be released "within this year." Targeting the "professional-level AGI" for the enterprise market Suleiman said that Microsoft Corporation's goal is to seize more market share in the enterprise market by creating a "professional-level general artificial intelligence." These AI tools can automate daily tasks for knowledge workers such as lawyers, accountants, project managers, and marketers. "In the next 12 to 18 months, most of the tasks that white-collar workers do in front of computers will be fully automated by AI," he said. Within the next two to three years, AI agents will be able to work more efficiently within the workflows of large organizations. These AI tools are expected to have the ability to continuously learn, evolve, and autonomously perform more tasks. "Creating new models will become as popular as making podcasts or writing blogs," Suleiman said. "In the future, every organization and every person on Earth can design AI that meets their needs." Intensifying competition in the enterprise market, capital expenditure under pressure However, Microsoft Corporation has its competitors in the enterprise market. Anthropic has established a clear first-mover advantage in the field of AI programming tools, while OpenAI and Alphabet Inc. Class C are actively competing for enterprise-level AI contracts. Microsoft Corporation is expected to have capital expenditures of up to $140 billion in the current fiscal year (ending in June), primarily to invest in AI infrastructure. But investors are concerned that such large expenditures are fueling an AI "bubble" and dragging down the performance of large tech stocks. In the past month, Microsoft Corporation's stock price has fallen by about 13%. "There is no doubt that we are in an unprecedented era, and the market is attempting to figure out how the next five years will unfold," Suleiman admitted. "But everyone is confident that these investments will eventually revenue and profits." Medical "super intelligence" and the vision of "human-centered" AI Suleiman said that another key focus for Microsoft Corporation is to drive the application of AI in the healthcare sector, with the goal of building a "medical super intelligence" - using AI to alleviate challenges such as shortages of medical personnel and long waiting times in the healthcare system. Microsoft Corporation launched an AI-assisted diagnostic tool last year, which reportedly outperforms human doctors in some tasks. He also emphasized that Microsoft Corporation pursues a "human-centered super intelligence" - AI technology that always remains under human control, in response to concerns in the industry about the development of "difficult-to-regulate" advanced AI systems by some companies. "We must establish guidelines: only when a system is truly controllable and able to operate in a manner that serves humans should it be introduced into the real world," Suleiman said. "These tools, like any other technology in the past, are aimed at enhancing human well-being and serving humans, not surpassing them."