Mustafa Suleyman Declares Microsoft 'Set Free' From OpenAI to Pursue Superintelligence

Microsoft’s artificial intelligence chief says the company has been “set free” from its long reliance on OpenAI, positioning Microsoft to push more aggressively toward its own vision of superintelligence, according to VentureBeat. The remarks from Mustafa Suleyman mark a shift in how Microsoft publicly frames its AI strategy after years in which its Copilot products and broader AI ambitions were closely tied to OpenAI’s models.

For much of the last three years, Microsoft’s AI identity was effectively intertwined with OpenAI, whose technology helped drive the rapid expansion of Microsoft’s AI offerings across consumer and enterprise products, VentureBeat reported. That partnership, backed by a cumulative investment of more than $13 billion, gave Microsoft early access to leading models and helped lift its market value, but it also made OpenAI central to Microsoft’s public AI story.

Suleyman’s comments suggest Microsoft now wants to be seen as building a more independent AI stack, with its own models, products and long-term research direction. In VentureBeat’s account, he argued that the company’s reduced dependence on OpenAI gives it room to pursue “superintelligence,” a term usually used to describe AI systems that exceed human capability across a wide range of tasks.

The shift comes as Microsoft continues to push Copilot deeper into the enterprise market, where the company says the main challenge is no longer just producing capable AI but managing it safely and effectively. At Build 2026, Microsoft emphasized that successful AI agents need reliable context, governance, identity, memory and secure access to company data, according to VentureBeat.

That message lines up with remarks from chief executive Satya Nadella, who said AI agents should be treated like employees, with identities, permissions and audits, according to Business Insider. The idea reflects a broader push inside Microsoft to make AI systems behave more like managed workers inside corporate environments rather than free-floating chatbots.

The business case is significant because enterprises are increasingly deploying AI agents to handle work across teams and systems, but researchers and vendors have warned that these tools often do not share what they learn from one user to another. VentureBeat reported that when one person corrects an AI agent, the improvement often does not transfer to colleagues using the same tool, which limits productivity gains unless companies build shared memory and governance layers.

Taken together, Microsoft’s latest remarks point to a company trying to move from partner to platform owner in AI. The immediate stakes are commercial, since enterprise adoption is where Microsoft expects much of its AI revenue growth, but the larger implication is strategic: Microsoft is signaling that its next phase will be defined less by OpenAI’s models and more by its own effort to build and control the infrastructure for advanced AI systems.