GOOGMarch 19, 2026 at 1:46 PM UTCSoftware & Services

Google Secures Power Deals to Support AI Data Center Expansion Amid Rising Infrastructure Costs

Read source article

What happened

Google has signed agreements with five U.S. utilities to reduce electricity use during peak demand, aiming to secure power for its fast-growing data centers. This move comes as Alphabet faces significant capital expenditure for AI infrastructure, with FY2026 capex guided at $175–185B and technical infrastructure costs expected to rise sharply. The DeepValue report highlights compute and power constraints as key headwinds, noting that energy costs are part of the increasing operational expenses. While these deals may help manage near-term power availability risks, they do not address the broader capex commitments or the critical need for Cloud backlog conversion. Alphabet's investment thesis remains centered on proving AI-driven margin stability and demand visibility, which this news does not materially alter.

Implication

The utility agreements could slightly mitigate energy cost volatility, a component of the rising technical infrastructure expenses flagged in the report. However, they are incremental compared to the $149.1B in purchase commitments and $42.6B in data center leases, which limit flexibility and underscore capex intensity. Investors should view this as a tactical move to address supply constraints, not a strategic change that reduces the need for evidence on Cloud backlog near $240B. The core risks—AI Search monetization shifts and margin compression from depreciation—remain unchanged, reinforcing the WAIT rating. Until quarterly data confirms backlog stability and margins above 31%, the stock's valuation at 28.1x P/E still prices in optimistic assumptions about capex payoffs.

Thesis delta

The news slightly offsets energy-related headwinds by demonstrating proactive cost management, but it does not alter the fundamental thesis. No shift in the investment call is warranted, as the key proofs—Cloud backlog conversion and AI Search monetization stability—remain unproven and critical for risk-adjusted returns.

Confidence

high