Meta and Microsoft have turned the AI race into a real estate race, with future data center lease commitments across Microsoft, Meta, Oracle, Amazon, Google, and CoreWeave climbing past $700 billion. Microsoft carries about $155 billion in total future lease obligations, while Meta carries about $104 billion. In one quarter, Meta and Microsoft each added roughly $50 billion in fresh data center lease commitments. The hyperscalers are locking up capacity for anticipated AI demand. Wholesale data center operators and REITs like Digital Realty and Equinix are benefiting, as landlords with powered space now hold leverage. Power is the choke point, as AI racks draw far more power than ordinary cloud workloads. Spending estimates vary: Axios reported in February 2026 that hyperscaler spending was expected to reach about $610 billion this year at the midpoint of company guidance, while a later Financial Times compilation put planned 2026 capex for Google, Microsoft, Meta, and Amazon at roughly $725 billion.