OpenAI has completed its controversial for-profit restructuring, establishing a new public benefit corporation called OpenAI Group PBC, while its original nonprofit entity is now the OpenAI Foundation. The nonprofit will hold equity in the for-profit arm, currently valued at approximately $130 billion, and will focus its initial $25 billion investment on healthcare, disease research, and AI resilience. It will also receive additional ownership stakes if the for-profit reaches certain valuation milestones. This restructuring, finalized after over a year of negotiations with California and Delaware attorneys general, marks a pivotal shift from OpenAI’s original plan, which would have stripped the nonprofit of all control. Instead, the new structure preserves the nonprofit’s oversight role and equity stake, up to $100 billion. The deal comes amid a long legal battle with Elon Musk, who co-founded OpenAI in 2015 as a nonprofit and has sued both the company and CEO Sam Altman to block the conversion. Musk’s opposition centered on concerns that the nonprofit’s mission was being compromised. The new structure, however, has been approved, allowing OpenAI to move forward and avoid losing up to $10 billion in promised SoftBank investment. A key part of the restructuring is a revised agreement with Microsoft, which has reduced its ownership stake in the for-profit from 32.5% to roughly 27%, now valued at about $135 billion. Crucially, Microsoft’s IP rights are now extended through 2032 and include access to models and products post-AGI (artificial general intelligence), provided safety guardrails are in place. This resolves a major sticking point: the original “AGI clause,” which allowed Microsoft to lose access to OpenAI’s technology once AGI was achieved. Under the new deal, AGI declarations will be verified by an independent expert panel, not just OpenAI alone, adding transparency and reducing unilateral control. Microsoft can now also pursue AGI independently or with third parties, including competitors like Anthropic, with whom it has already begun sourcing AI. This marks a significant shift from the previous exclusive partnership. However, Microsoft’s IP rights do not extend to OpenAI’s upcoming consumer hardware, a device being developed with Apple’s famed designer Jony Ive. This suggests OpenAI is betting heavily on personal AI devices, with Altman describing the goal as creating a “personal AGI” that assists users in daily life. The new agreement also allows OpenAI to collaborate with third parties, release open-weight models, and no longer requires Microsoft to have first refusal on compute services. OpenAI remains contractually obligated to purchase an additional $250 billion in Azure cloud services from Microsoft. Despite these advances, questions remain. The identity and selection process for the AGI verification panel have not been disclosed. Altman acknowledged that the term AGI has become “hugely overloaded,” and instead of chasing a vague definition, OpenAI now aims to achieve a fully automated AI researcher by March 2028—a more tangible milestone. The stakes are immense. AGI represents a potential paradigm shift in technology, economics, and society. With Microsoft now empowered to pursue AGI independently using OpenAI’s IP, the race has intensified. Yet OpenAI insists its mission remains focused on safety and long-term benefit. As Altman stated, AGI will be a multi-year process, not a single event. The new deal, while a win for OpenAI’s growth and autonomy, also underscores the complex, high-stakes balance between innovation, control, and responsibility in the AI era.