Satya Nadella, the CEO of Microsoft, recently emphasized at the World Economic Forum in Davos that the key to AI sovereignty lies not in where AI processes run but rather in who controls them. Nadella argued that corporate AI sovereignty depends on the ability to manage models trained on proprietary knowledge, rendering the physical location of infrastructure as secondary in importance. He stated, \”In the AI age, sovereignty comes from embedding a company’s tacit knowledge in models you control.\” This perspective comes as Microsoft navigates global data sovereignty challenges, notably within the EU, where complete protection from US access demands has been an issue. Nadella suggests the conversation should shift towards sovereignty in AI, emphasizing encryption’s role in addressing concerns while noting the constraints of light speed on datacenter placements. A recent World Economic Forum paper outlines AI sovereignty as economies shaping AI ecosystems aligned with their values while maintaining control and resilience through local investment and international collaboration. Despite challenges, Nadella believes that true corporate and national advantages—preserved in the AI era—will define real sovereignty. His remarks highlight Microsoft’s extensive commitment to AI, demonstrated by significant investments in OpenAI and integration of AI into its portfolio.