Microsoft has quietly restructured its AI division around one goal: building superintelligence, and selling it to every business on the planet.
Mustafa Suleyman doesn't waste time on small ambitions. As Microsoft's first-ever CEO of AI, he has spent the better part of a year laying the groundwork for what comes after chatbots and copilots. Now, after a significant corporate reshuffling in mid-March, his mission has a name and a clear direction. Microsoft is going after superintelligence, and it intends to own the commercial layer of that technology.
As The Verge recently reported, Suleyman had been preparing for this pivot for roughly nine months before the company made the news public last month. The catalyst was a renegotiated contract with OpenAI, the San Francisco-based lab behind ChatGPT and GPT-4. That new arrangement, Suleyman explained, is what formally "unlocked" Microsoft's ability to pursue superintelligence on its own terms. But the strategic blueprint existed long before the legal documents were signed.
This matters because it signals a fundamental shift in how the world's most valuable companies are thinking about artificial intelligence. The first wave of the AI boom was about catching up. Everyone rushed to embed large language models into search engines, productivity suites, and customer service platforms. The second wave, the one Microsoft is now positioning for, is about building systems that surpass human cognitive capabilities across virtually every economically valuable domain. That is the working definition of superintelligence, and the race to commercialize it is already underway.
Microsoft's relationship with OpenAI has always been complicated. The Redmond-based tech giant has invested more than $13 billion into OpenAI since 2019, securing exclusive cloud computing rights and deep integration with Azure. But that partnership also came with constraints. If OpenAI achieved certain thresholds of capability, governance provisions could shift. Microsoft needed clarity about what it could build independently, and on what timeline.
The renegotiated contract gave Microsoft the freedom to develop its own superintelligence infrastructure without stepping on OpenAI's toes or creating internal conflicts about where research dollars and compute resources should flow. For Suleyman, who previously co-founded DeepMind before it was acquired by Google and later launched the AI startup Inflection AI, this was the green light he needed. His team can now pursue ambitious research while also building the enterprise products that will turn breakthroughs into revenue.
And enterprise revenue is the real prize here. While consumers play with ChatGPT and debate whether AI-generated art is a novelty or a threat, the businesses buying AI services are spending real money. Enterprise AI spending is projected to exceed $500 billion by 2027, according to estimates from IDC, and Microsoft wants the largest possible share. Superintelligence isn't just a research trophy. It's the foundation for systems that can design drugs, optimize global supply chains, write and debug software autonomously, and manage complex financial portfolios with minimal human oversight.
The Competitive Landscape Is Reshaping Fast
Microsoft isn't alone in this pursuit. Google DeepMind, where Suleyman once worked, is pursuing similar capabilities under Demis Hassabis. Anthropic, the maker of the Claude AI assistant, has framed its entire mission around building safe, highly capable AI systems. Meta continues to invest in open-source AI infrastructure, hoping that widespread adoption of its Llama models will give it leverage in the next generation of AI tooling. And Amazon is pouring resources into its Bedrock platform and its partnership with AI startup Anthropic, ensuring AWS customers have access to the most powerful models available.
What differentiates Microsoft's approach is the explicit focus on the business wrapper around superintelligence. Suleyman's restructured division isn't just a research lab. It's designed to move from fundamental breakthroughs to commercial products on a compressed timeline. The March restructuring, which handed off some of his earlier operational duties, was about giving him the bandwidth to focus exclusively on this long-term bet. That kind of organizational commitment tells you something about where Microsoft sees the puck heading.
For startups and enterprise leaders watching this unfold, the practical takeaway is straightforward. The AI tools you're integrating today are the floor, not the ceiling. Within a few years, the competitive advantage will shift from who has access to a capable language model to who can most effectively deploy systems that reason, plan, and execute at a level that matches or exceeds top-tier human experts. Microsoft is betting its future on that transition, and the infrastructure decisions being made now will determine which companies are positioned to benefit and which ones get left behind.
