In a move that has sent shockwaves through the tech industry and Wall Street, Microsoft AI CEO Mustafa Suleyman has effectively confirmed that the software giant is aggressively “dumping” its heavy reliance on OpenAI in favor of “true AI self-sufficiency.” Speaking in a recent interview, Suleyman revealed that his personal mission at Microsoft is to build a world-class superintelligence, signaling a strategic decoupling from the startup that has powered Microsoft’s AI revolution for years. This shift follows the October 2025 restructuring of their partnership, which granted Microsoft the right to independently pursue Artificial General Intelligence (AGI) and reduced its profit-sharing role to a 27% ownership stake. Suleyman emphasized that Microsoft is now prioritizing the development of its own frontier foundation models, such as the recently previewed MAI-1, leveraging its massive “gigawatt-scale” computing power to create in-house systems that could eventually replace OpenAI models across the Copilot and Office ecosystems.
Beyond the corporate divorce, Suleyman issued a stark warning regarding the future of the global workforce, predicting that most white-collar tasks—including those performed by lawyers, accountants, and project managers—could be “fully automated” by AI within the next 12 to 18 months. He described a future where creating a custom AI model will be as simple as launching a podcast or a blog, potentially leading to what some are calling an “economic earthquake.” While Microsoft continues to host various models from Meta, Mistral, and even Anthropic in its data centers, the push toward internal “superintelligence” marks a definitive end to the era of exclusive dependency. Investors have reacted with caution to the news, as concerns over the durability of Microsoft’s sales backlog—previously heavily tied to OpenAI’s success—triggered a historic single-day wipeout of $357 billion in market value earlier this week.