Microsoft's AI chief has stated that the company is pursuing "true self-sufficiency" in artificial intelligence by developing its own powerful models and decreasing its dependence on OpenAI.

Mustafa Suleyman explained in an interview that this strategic shift followed a restructuring of the relationship between the two companies last October. This move has prompted the $3 trillion company to begin independently building its most advanced technologies, rather than relying on external partners.

The Google DeepMind co-founder, who joined Microsoft in 2024, stated: "We must develop our own cutting-edge foundation models, equipped with gigawatt-scale computing power and some of the world's top AI training teams."

The software giant is making significant investments in assembling and organizing the massive datasets required to train advanced systems. He added: "This is our true self-sufficiency mission."

Microsoft, one of OpenAI's earliest and largest investors, has long depended on OpenAI's models to power its own AI tools, such as the Copilot software assistant.

Last year, Microsoft agreed to allow the startup to complete its corporate restructuring while retaining its $135 billion stake in the ChatGPT developer and securing access to its most advanced models until 2032.

However, the agreement also grants OpenAI greater freedom to seek new investors and infrastructure partners, potentially transforming it into a more direct competitor.

Microsoft is also diversifying its bets by investing in other model developers like Anthropic and Mistral. Simultaneously, the company is accelerating the development of its proprietary models. Suleyman indicated that these internal models would be released "sometime this year."

Suleyman specifically noted that Microsoft's goal is to capture more enterprise market share by advancing "professional-grade AGI" (Artificial General Intelligence), meaning developing powerful AI tools capable of performing routine tasks for knowledge workers.

He said: "White-collar work—whether it's lawyers, accountants, project managers, or marketers sitting at computers—will see the majority of its tasks fully automated by AI within the next 12 to 18 months."

He further added that within the next two to three years, these AI agents will achieve better coordination within the workflows of large organizations. These AI tools will also possess continuous learning and improvement capabilities and will be able to take more autonomous actions.

He stated: "Creating a new model will become as simple as producing a podcast or writing a blog. In the future, it will be possible to design AI tailored to the needs of every organization and individual on the planet."

Nevertheless, Microsoft faces intense competition in the enterprise market. Anthropic has taken a lead in AI programming tools, while OpenAI and Google are also heavily investing to compete for lucrative enterprise AI contracts.

Microsoft anticipates capital expenditures to reach $140 billion in the fiscal year ending in June, allocated towards expanding its AI infrastructure.

Investors are concerned that such spending is inflating an AI "bubble," which has pressured large tech stocks. Microsoft's share price has fallen more than 13% over the past month.

Suleyman commented: "There's no question, these are unprecedented times, and I believe the market is trying to understand the development path for the next five years." But he added: "We have no doubt that these returns will ultimately compound and be reflected in revenue and profits."

Suleyman also mentioned that another key focus for Microsoft is applying AI to healthcare, aiming to build a "medical super-intelligence" that uses AI programs to address shortages of healthcare workers and long wait times in medical systems. Last year, the company launched an AI diagnostic tool, claiming it outperforms doctors in certain tasks.

He added that Microsoft's objective is to create a "humanistic super-intelligence"—AI technology that remains under human control—addressing concerns about some AI labs developing powerful technologies too rapidly, potentially beyond their creators' oversight.

He stated: "We must reset this premise, only bringing systems into the world that we are confident we can control and that serve us in a subordinate manner."

"These tools, like any technology before them, are designed to enhance human well-being and serve humanity, not to surpass it."