TLDRs;
- Microsoft forms MAI Superintelligence Team led by Mustafa Suleyman to build advanced, human-centered AI systems.
- Revised OpenAI deal gives Microsoft freedom to train its own large models for the first time.
- The company retains exclusive Azure rights and commits up to $250B through 2032 under the renewed partnership.
- MAI efforts could supercharge cloud, compute, and chip sectors amid surging AI infrastructure investments.
Microsoft has officially launched the MAI Superintelligence Team, a new research initiative dedicated to building what CEO Satya Nadella describes as “humanist superintelligence” , AI systems designed to benefit humanity while maintaining alignment with human values.
The team will be led by Mustafa Suleyman, the co-founder of both Google DeepMind and Inflection AI, marking one of Microsoft’s most ambitious moves yet in artificial general intelligence (AGI) development.
The creation of MAI comes on the heels of a revised agreement with OpenAI, which had previously restricted Microsoft’s ability to independently train large-scale models. Those limitations have now been lifted, enabling the company to pursue its own AGI research while maintaining deep commercial and technological ties with OpenAI through at least 2030.
‘Humanist Superintelligence’ Vision Takes Shape
Suleyman’s new team embodies a guiding principle he has long championed: AI systems that serve people first. Under this vision, Microsoft’s next-generation AI would integrate ethics, transparency, and safety as core components, rather than afterthoughts.
Joining Suleyman is Karén Simonyan, who will serve as the group’s chief scientist. Simonyan, a fellow Inflection alumnus and former DeepMind researcher, brings a wealth of experience in neural architectures and model scaling. The team also includes talent drawn from Meta, Anthropic, Google DeepMind, and OpenAI, reflecting a cross-industry push toward responsible AGI research.
The focus on “humanist” AI suggests a deliberate contrast to the race for raw intelligence capabilities seen elsewhere in the industry.
“Our mission,” Suleyman reportedly told Fortune, “is to ensure that superintelligence reflects the best of humanity, not just its efficiency or competitiveness.”
Revised OpenAI Pact Unlocks Research Freedom
The amended OpenAI deal is central to Microsoft’s newfound autonomy. According to UBS analysis, the agreement restores Microsoft’s right to train large language models independently, while preserving exclusive intellectual property (IP) and Azure API exclusivity until AGI is formally achieved.
The pact also extends OpenAI hosting and IP rights through 2032, ensuring continuity in joint deployments like Copilot and Azure AI services. UBS estimates Microsoft’s total commitment to the partnership at $250 billion over seven to eight years, a reflection of the company’s long-term bet on AGI as a cornerstone of its future business model.
Crucially, the MAI initiative operates in parallel with Microsoft’s existing OpenAI-backed efforts, creating what analysts have called a “dual-track AGI strategy.”
Massive Infrastructure and Compute Implications
This dual-track model could have ripple effects far beyond software. UBS projects Microsoft’s annual OpenAI-related expenditures at $30–$35 billion, much of which fuels demand for cloud compute and high-performance chips.
The company is already building new AI datacenters, including a massive Wisconsin site equipped with hundreds of thousands of Nvidia Blackwell GB200 GPUs. A second facility, expected to come online by 2027, could push total power needs past 900 megawatts, underscoring the colossal scale of Microsoft’s AI ambitions.
Meanwhile, the company is also developing custom silicon to reduce reliance on Nvidia, opening the door for emerging chipmakers to compete for MAI workloads.
Get 3 Free Stock Ebooks
Discover top-performing stocks in AI, Crypto, and Technology with expert analysis.
- Top 10 AI Stocks - Leading AI companies
- Top 10 Crypto Stocks - Blockchain leaders
- Top 10 Tech Stocks - Tech giants
