Big Tech's AI Spending Is on Track to Top $700 Billion This Year. Here's Who May Cash In Next.
The article says major tech firms are set to spend over $700 billion on AI-related data centers and chips in 2026, citing planned capex of about $200 billion (Amazon), $190 billion (Microsoft), up to $190 billion (Alphabet), and $125–$145 billion (Meta). It highlights American Electric Power (AEP), which in Q1 2026 added 7 GW of future load, raising contracted load by 2030 to 63 GW, nearly 90% data centers. AEP raised its five-year capital plan to $78 billion, reported Q1 revenue up ~10% to $6.0
AI-driven contracted load growth is a near-to-mid term earnings and rate-base support story for AEP, but execution and funding/demand risks remain.
AEP reports rising contracted data-center load to 63 GW by 2030 and raised its five-year capex plan to $78B.
Bias modestly positive with pullbacks possible on funding/regulatory or AI-spend slowdown concerns.
Background
The piece frames Big Tech’s 2026 AI capex surge as flowing through to electricity transmission/distribution utilities via contracted data-center load.
Why it matters
For AEP, the key transmission/load additions and capex/rate-base growth expectations are the core bullish mechanism; the main bearish mechanisms are build delays, dilution from equity issuance, and potential AI demand downdraft.
Market relevance
Traders can use the article’s specific AEP load/capex updates and valuation context (forward P/E and dividend) to gauge whether the market is already pricing the AI infrastructure ramp.
Market effects
Reinforces the AI data-center power demand theme as a fundamental driver for regulated utilities’ rate-base growth.
Highlights load concentration in Indiana, Ohio, Oklahoma, and Texas, implying heavier grid buildout needs in those states.
Primarily US utility demand; global relevance is indirect via hyperscaler AI capex cycle affecting US power infrastructure.
Alternative perspectives
If hyperscaler AI capex slows materially, contracted load may not translate into realized utilization, leaving AEP with stranded/overbuilt assets despite take-or-pay protections.
Regulatory approval timing and grid-operator interconnection bottlenecks could delay revenue recognition and increase financing costs, offsetting the growth narrative.
Key entities
- companyAmerican Electric Power
Utility highlighted as one of the most exposed beneficiaries of AI-driven data-center electricity demand, with updated contracted load and capex plans.



