AI Data Centers Are Pushing Electricity Prices Higher — Energy Teams Are Responding

The electricity bill for running artificial intelligence is coming due — and it's landing on ratepayers and energy-intensive businesses alike.

According to CNBC, the rapid expansion of AI data centers is igniting a debate over who pays for the infrastructure needed to power them. Tech companies are securing gigawatt-scale power purchase agreements, straining regional grids and pushing wholesale prices higher. The U.S. Energy Information Administration now projects average wholesale electricity prices to reach $51/MWh in 2026 — an 8.5% increase year-over-year — with AI demand cited as a primary structural driver, according to Utility Dive.

The price impact isn't uniform. Regions with heavy data center concentrations — Northern Virginia, central Texas, the Pacific Northwest — are seeing spot price volatility well above national averages. Energy procurement teams and industrial energy consumers that were managing against stable price curves are now navigating a market that looks more like financial trading than utilities.

S&P Global's Response: Acquiring AI for the Energy Market

The market intelligence sector is betting on data and automation to navigate the new volatility. On March 18, 2026, S&P Global announced the acquisition of Enertel AI Corporation to enhance its power market analytics platform. Enertel AI's technology uses graph neural networks (GNNs) to deliver real-time, node-level electricity price forecasting across transmission networks.

The acquisition signals where the energy industry is heading: away from backward-looking price reports and toward predictive, API-accessible market intelligence that teams can integrate directly into procurement workflows.

"The convergence of AI-driven demand forecasting and real-time market data is no longer a nice-to-have," said an analyst cited in the deal announcement. "For commercial and industrial energy buyers, it's becoming table stakes for competitive procurement."

The Hidden Costs for Businesses

Rising electricity prices create compounding pressure across sectors:

Manufacturing and logistics: Energy typically represents 15-30% of operating costs for heavy manufacturers. A sustained 8-10% annual price increase materially affects margins for facilities not covered by long-term power purchase agreements.

Data center operators: Ironically, the same AI infrastructure that is driving price increases makes existing data centers more expensive to run. Operators without flexible load management are seeing power costs outpace efficiency gains.

Retailers and commercial real estate: Large building portfolios face rising utility costs that are difficult to pass through to tenants on fixed leases.

For all of these groups, the ability to monitor real-time spot prices, track regional volatility indices, and forecast short-term price movements has moved from "useful analytics" to a procurement necessity.

The Forecasting Challenge

Traditional energy price forecasting relied on historical patterns and seasonal adjustments — models that assumed a relatively stable supply/demand equilibrium. That equilibrium has been disrupted by:

1. Variable renewable generation — Wind and solar create high-frequency price swings that historical averages don't capture
2. AI demand spikes — Data center load is increasingly variable and correlated with compute-intensive AI workloads
3. Transmission constraints — Grid infrastructure hasn't kept pace with demand growth, creating regional price separation

Effective energy management in this environment requires continuous data access, not quarterly reports.

What Energy Teams Are Actually Doing

Procurement teams responding to the new volatility environment are adopting several approaches:

Real-time spot price monitoring: Setting automated alerts when prices in specific markets cross thresholds that affect hedging decisions or operational flexibility.

Volatility index tracking: Monitoring not just price levels but price uncertainty — high-volatility periods warrant different hedging strategies than stable markets.

7-14 day price forecasting: Short-term forecasts inform decisions about curtailment, flexible load shifting, and spot vs. contract procurement timing.

Multi-market comparison: Large industrial consumers with operations across regions can optimize load distribution in real time based on relative regional prices.

Automating Energy Market Intelligence with APIs

The Energy Volatility API provides developers and energy analysts with programmatic access to real-time spot prices, historical data, and short-term forecasting for major US energy markets including ERCOT, PJM, CAISO, and MISO.

Rather than manually pulling data from multiple ISO websites or waiting for daily reports, teams can integrate real-time market data directly into:

• Internal dashboards and BI tools
• Procurement approval workflows
• Automated hedging systems
• ERP and energy management platforms

A typical integration requires less than a day of development work using the REST API and can be embedded in existing tools without building or maintaining energy data infrastructure.

Looking Ahead

According to Axios, the 2026 energy market will be defined by the tension between accelerating AI-driven demand and the slower pace of grid modernization and new generation capacity. Price volatility is expected to remain elevated through at least 2028 as data center construction outpaces transmission upgrades in high-demand corridors.

For organizations with significant energy costs, the question is no longer whether to invest in energy market intelligence — it's how quickly they can automate it.

Sources

• Who is really footing the AI energy bill? — CNBC, March 13, 2026
• S&P Global Acquires Enertel AI Corporation to Enhance Power Market Offering — S&P Global Press Release, March 18, 2026
• Electricity prices to continue rise in 2026: EIA — Utility Dive
• Electricity prices, AI and the "climate reset": What to watch for in 2026 — Axios, January 6, 2026