The AI arms race has officially moved from the cloud to the ground. As 2025 draws to a close, two major developments this week underscore a fundamental shift in how technology giants are competing: Alphabet's $4.75 billion acquisition of Intersect and ByteDance's reported $23 billion capital expenditure plan for 2026. The message is clear: the next phase of AI dominance will be won by those who control power, real estate, and silicon—not just algorithms.
What Happened
On December 22, Alphabet announced a definitive agreement to acquire Intersect, a data center and energy infrastructure provider, for $4.75 billion. The deal, expected to close in the first half of 2026, brings multiple gigawatts of energy and data projects already in development under Google's control, including a co-located data center and power site under construction in Texas.
The strategic rationale is straightforward: AI workloads are power-hungry, and Alphabet needs to ensure "compute and power arrive together" to eliminate capacity bottlenecks that have constrained its Gemini AI ambitions.
Meanwhile, TikTok parent ByteDance has made preliminary plans for capital expenditure of 160 billion yuan ($22.74 billion) in 2026, specifically targeting AI infrastructure. Of this amount, approximately 85 billion yuan has been budgeted for AI processors alone—a significant commitment given ongoing uncertainty over access to Nvidia chips due to U.S. export restrictions.
By the Numbers
- $4.75 billion – Alphabet's acquisition price for Intersect
- $23 billion (160 billion yuan) – ByteDance's planned 2026 capex for AI infrastructure
- $61 billion – Record data center deal volume in 2025, according to CNBC
- $300 billion+ – Combined 2025 capex from Amazon, Microsoft, Alphabet, and Meta, per Lucidity Insights
- 85 billion yuan (~$12 billion) – ByteDance's budget specifically for AI processors in 2026
Market Reaction
Investors have largely embraced Big Tech's infrastructure pivot, viewing it as a necessary moat-building exercise rather than reckless spending. The Intersect acquisition positions Alphabet to compete more effectively against Microsoft's $80 billion data center commitment and Amazon's industry-leading $100 billion capex plan for 2025.
ByteDance's aggressive spending signals that the AI infrastructure race is now truly global. However, analysts note the geopolitical complexities: with U.S. export controls limiting access to cutting-edge Nvidia GPUs, ByteDance may need to rely more heavily on domestic Chinese chip suppliers or alternative architectures, potentially affecting performance parity with Western competitors.
The broader data center M&A market reflects sustained investor appetite. According to CNBC, data center deals hit a record $61 billion in 2025, even as some investors express concerns about the sustainability of AI-driven valuations.
The Bigger Picture
These developments represent a structural transformation in how technology companies operate. The hyperscalers are no longer purely software companies—they are becoming energy and real estate conglomerates.
Alphabet's Intersect acquisition is particularly telling. Rather than relying on traditional utilities and third-party energy developers, Google is internalizing energy development to eliminate the mismatch between compute capacity and power availability. The deal includes exploration of emerging technologies to increase and diversify energy supply, including carbon capture initiatives.
The scale of investment is staggering. Amazon leads with $100 billion in 2025 capex, followed by Microsoft at $80 billion, Google at $75 billion, and Meta at $60-65 billion. ByteDance's $23 billion 2026 commitment would place it firmly in the same league as Western giants—a remarkable position for a company that didn't exist a decade ago.
Global cloud infrastructure spending reached $102.6 billion in Q3 2025 alone, up 25% year-over-year, according to Omdia research. This spending cycle represents the most aggressive tech investment since the mobile boom.
What to Watch
Alphabet's integration timeline: The Intersect deal is expected to close in H1 2026. Watch for updates on how quickly the acquired energy projects come online and whether they meaningfully accelerate Google's AI compute capacity.
ByteDance's chip sourcing: With 85 billion yuan earmarked for AI processors, ByteDance's ability to secure adequate supply amid U.S. export restrictions will be a key indicator of whether Chinese tech can maintain AI competitiveness.
Energy bottlenecks: As data center power consumption surges, watch for regulatory pushback, grid capacity constraints, and further energy M&A activity across the sector.
Return on infrastructure investment: Microsoft's Azure has seen measurable lift from AI demand, with 31% average quarterly growth in 2024. Investors will be scrutinizing whether these massive capex commitments translate into revenue growth or become stranded assets.
The infrastructure wars have begun. In this new phase of AI competition, the winners won't just have the best models—they'll have the power plants, the data centers, and the supply chains to run them at scale.