Duke Energy unveiled an industry-record $103 billion five-year capital plan that has now become the defining benchmark for how American utilities intend to fund the artificial intelligence buildout. Reaffirmed in a Fortune profile on April 25, 2026, the plan dwarfs every regulated utility capex program ever filed and signals that the data center electricity surge has moved from a forecasting exercise into hard concrete and steel commitments. With 14 GW of new generation, 4.5 GW of batteries, and a quiet pivot toward small modular reactors at Belews Creek, Duke Energy is rewriting the rulebook for utility growth in the AI era.

The numbers are staggering even by recent infrastructure standards. Duke is committing more than $1 billion in capital every month, planning $10 billion in new equity issuance between 2027 and 2030, and projecting 9.6% earnings-based growth through the end of the decade. Yet the announcement landed in the same week that critics filed regulatory complaints accusing Duke of socializing data center costs onto residential ratepayers, setting up the most consequential utility rate fight of the AI cycle.

The $103 Billion Headline: What Duke Energy Just Promised Wall Street

Duke Energy first disclosed the $103 billion 2026-2030 capital plan on its February 11, 2026 fourth-quarter earnings call, then doubled down with a sweeping affordability narrative reiterated in a high-profile Fortune feature on April 25, 2026. The five-year program represents a $16 billion increase – roughly 18% – over the company’s prior $87 billion plan, and CEO Harry Sideris described it as “the largest fully regulated capital plan in the industry, focused on critical energy infrastructure investments that strengthen the system and serve increasing load.”

The size matters because Duke serves roughly 8.4 million customers across six states – North Carolina, South Carolina, Florida, Indiana, Ohio, and Kentucky – territories that have become epicenters of hyperscale data center development. The Carolinas alone now host announced or committed campuses from Amazon, Meta, Microsoft, and Google, and Indiana’s central location near Chicago fiber routes has made it one of the fastest-growing data center markets in the country.

Approximately 65% of the $103 billion flows directly to grid infrastructure and new power generation. The remainder funds nuclear license extensions, environmental compliance, transmission expansion, and the rate-base buildout needed to absorb 14 GW of incremental generation by 2030. Duke’s prior 2024-2028 plan totaled $73 billion, meaning the utility has effectively raised its build pace by more than $30 billion per five-year window in just two annual revisions.

Why April 25, 2026 Matters: Data Centers Take Center Stage

While the headline number first surfaced in February, the April 25, 2026 Fortune profile reframed the plan around a single word: affordability. That pivot matters because it confirms what utility analysts had been whispering for months – Duke now believes it must publicly defend its data center alignment to keep state commissions on side. Sideris told Fortune the company is “balancing the largest growth opportunity in our history with the obligation to keep bills as flat as we can,” signaling that Duke expects political turbulence well into 2027.

👁 Why April 25, 2026 Matters: Data Centers Take Center Stage

The timing dovetails with Fortune’s separate April 20, 2026 reporting that data centers drove roughly half of all U.S. electricity demand growth in 2025, a statistic that has reshaped how every regional transmission organization plans capacity. PJM, MISO, and SERC interconnection queues have ballooned past 2,000 GW of pending requests, and Duke’s territory has been the single largest source of new connection requests in the SERC footprint.

The April announcement also followed Duke’s 2026 Q1 trading update, which confirmed the company is “deploying more than $1 billion in capital every month, with acceleration expected in 2027 and 2028 as data centers connect,” according to investor materials referenced on the February call. That run-rate puts Duke on pace to outspend even the largest oil and gas supermajors on a quarterly capital basis through the back half of the decade.

Inside the Capex Split: Where the $103 Billion Actually Goes

Duke’s investor materials disclose category-level allocations that reveal a heavy tilt toward generation and transmission, with relatively modest spending on customer-side technology. The grid hardening line item alone – covering substations, distribution upgrades, storm resilience, and wildfire mitigation in the Carolinas – accounts for the largest share of any single bucket. New generation capacity, including 5 GW of natural gas combined-cycle plants, sits second.

Capex Category (2026-2030)	Allocation	Share of $103B	Primary Driver
Grid infrastructure and modernization	~$33B	32%	Data center interconnection, storm hardening
New natural gas generation	~$22B	21%	5 GW Carolinas + Indiana firming capacity
Renewables, solar, and storage	~$18B	17%	4.5 GW battery, utility solar additions
Transmission expansion	~$14B	14%	SERC and MISO interconnection upgrades
Nuclear (existing fleet + SMR planning)	~$8B	8%	License extensions, Belews Creek SMR site work
Customer experience and other	~$8B	8%	Metering, IT, regulatory compliance

The split reveals a company doubling down on rate-based hardware rather than software or demand-response programs. That is the structural choice that makes Duke’s plan so attractive to long-only utility investors: nearly every dollar above flows through traditional cost-of-service ratemaking, generating contractually predictable returns over 30-to-40-year asset lives. It is also the reason consumer advocates are sounding alarms – every dollar of rate base is also a dollar that customers pay back with a regulated equity return on top.

The 14 GW Generation Surge and Duke’s New Natural Gas Push

The plan adds approximately 14 GW of incremental generation capacity across the five-year window, the most aggressive build cycle Duke has attempted since the 1970s nuclear era. Of that total, 5 GW is new natural gas combined-cycle capacity, with construction starting in the Carolinas and Indiana in 2026 and 2027. The natural gas tilt has drawn sharp criticism from environmental groups, but Duke has argued repeatedly that gas firming capacity is the only way to reliably integrate the data center loads coming online in 2027 and 2028.

Sideris pointed to interconnection queue data showing that data center customers are demanding 1 GW per facility in some cases, equivalent to the entire annual peak load of a mid-sized American city. Without firming generation that can ramp on demand, Duke argues, the regional grid simply cannot serve such loads while maintaining reserve margins required by NERC reliability standards.

The remaining capacity additions break down into roughly 4 GW of utility-scale solar, 4.5 GW of battery storage, and a sliver of upgraded hydro and nuclear uprates. Notably absent is any new offshore wind commitment – a category Duke has effectively shelved after federal permitting headwinds in 2025 and rising offshore turbine costs.

The Belews Creek SMR Pivot: Duke’s Quiet Nuclear Bet

One of the most consequential elements of Duke’s new plan is buried in a single line item: an early site permit application filed in December 2025 for a small modular reactor at the Belews Creek Steam Station in Stokes County, North Carolina. Duke has not formally selected a reactor vendor, but utility analysts widely expect a competition between GE Hitachi’s BWRX-300 and Holtec’s SMR-300 to anchor the site, with a target commercial operation date in the early 2030s.

👁 The Belews Creek SMR Pivot: Duke's Quiet Nuclear Bet

The Belews Creek pivot is significant for three reasons. First, the existing coal plant at the site already has interconnection rights, water rights, and a trained workforce. Second, the site sits inside Duke’s data center recruitment corridor between Charlotte and the Research Triangle. Third, the early permit positions Duke to qualify for federal nuclear loan guarantees and Department of Energy cost-share programs that other utilities have struggled to access.

Duke’s SMR move tracks an industry-wide pivot toward nuclear-anchored data center campuses. Amazon’s $11 billion Talen Energy deal in Pennsylvania, Microsoft’s Three Mile Island restart with Constellation, and Google’s Kairos Power partnership all share a similar thesis: hyperscale AI loads need 24/7 carbon-free baseload, and only nuclear can credibly provide it at scale by the early 2030s.

Battery Storage Ramp: 4.5 GW by 2031

Battery storage is the unsung hero of Duke’s plan. The company is committing to 4.5 GW of incremental battery capacity by 2031, a roughly 9x increase over its current installed base. The batteries are sized primarily for four-hour duration to match peak data center demand windows and serve as a transitional asset while gas plants and the eventual SMR come online.

Duke’s storage strategy diverges sharply from California-style daily solar shifting. In the Southeast, batteries are being deployed primarily for capacity firming, ancillary services, and momentary load smoothing during data center startup spikes. The company has indicated it will source roughly half of its battery installations from domestic manufacturers to qualify for Inflation Reduction Act bonus tax credits, even though those incentives have been narrowed during the 2025-2026 budget reconciliation cycle.

The cumulative effect of the gas-plus-battery strategy is a flexibility-first generation portfolio that can credibly serve the 73% year-over-year increase in data center power requests Duke disclosed earlier this year. Investors have welcomed the approach as a pragmatic bridge while regulators evaluate the longer-term SMR pipeline.

Capex Comparison: How Duke Stacks Up Against NextEra, Dominion, and Southern

The $103 billion figure makes sense only when compared with the broader utility cohort. Every major U.S. investor-owned utility has raised its capital plan in the past 18 months in response to data center demand, but Duke now sits comfortably at the top of the table both in absolute dollars and in growth rate.

Utility	5-Year Capex Plan	Service Footprint	Implied Annual Run Rate
Duke Energy	$103B (2026-2030)	NC, SC, FL, IN, OH, KY	~$20.6B
NextEra Energy	~$75B (2026-2029)	FL, plus FPL, NEER renewables	~$18.8B
Southern Company	~$63B (2025-2029)	GA, AL, MS	~$12.6B
Dominion Energy	~$50B (2025-2029)	VA, NC, SC, OH	~$10.0B
American Electric Power	~$54B (2025-2029)	11 states across PJM, SPP, ERCOT	~$10.8B

Duke’s plan stands out not just for its scale but for its geographic concentration. While AEP and Southern spread their spending across larger footprints, Duke is funneling capital into a handful of high-growth corridors – the Charlotte-Raleigh data center belt, central Indiana, and the Tampa-Orlando Florida cluster. That concentration makes for cleaner regulatory recovery, but it also means a single adverse rate case could materially impair the plan’s economics.

Industry analysts at Morningstar and Edison Electric Institute have flagged that the cumulative capex commitments from the top 10 U.S. utilities now exceed $1.4 trillion through 2030, a number that aligns with our prior coverage of the broader grid overhaul. Duke alone now represents roughly 7% of that total, an extraordinary concentration for a single regulated utility.

The Affordability Battle: Why Customer Bills Are the Real Risk

The single biggest risk to Duke’s $103 billion plan is not engineering or financing – it is political. Consumer advocacy group NC WARN issued a sharp news release on April 21, 2026, alleging that Duke’s plan would add roughly $60 billion to the rate base in the Carolinas alone within four years, with most of the cost recovered from residential and small-business customers. The group argues that data center customers should sign tariff structures that fully insulate other ratepayers from the cost of new generation built specifically to serve them.

👁 The Affordability Battle: Why Customer Bills Are the Real Risk

The political stakes have escalated quickly. North Carolina, South Carolina, and Indiana legislators are all weighing variations of “data center tariff” legislation that would require hyperscalers to pre-commit to long-term capacity contracts with minimum bill obligations. Duke has publicly supported tariff frameworks but has resisted proposals that would shift all stranded-asset risk to data center customers, arguing such terms would push hyperscalers to neighboring states.

“This is the most important utility regulatory cycle since deregulation,” said Travis Miller, energy strategist at Morningstar, in commentary published in February 2026. “Whoever sets the precedent for how data center load is allocated will define utility economics for the next twenty years.” Duke’s defense, articulated by Sideris on multiple investor calls, is that data center growth lifts state economies and ultimately spreads fixed costs across a larger sales base, lowering bills for everyone in the long run.

Wall Street’s Reaction: DUK Stock, Market Cap, and Valuation

Investors have responded to the plan with cautious enthusiasm. Duke Energy (NYSE: DUK) closed at $128.04 on April 23, 2026, with a 52-week range of $111.22 to $134.49, giving the company a market capitalization of approximately $101 billion on 778 million shares outstanding. The stock has roughly tracked the broader Utilities Select Sector SPDR (XLU) since the February announcement, gaining ground on dovish Fed expectations while occasionally lagging on rate-case headlines.

The valuation conversation centers on two competing narratives. Bulls argue that Duke’s 9.6% earnings-based growth target – anchored by the largest fully regulated rate base expansion in the industry – justifies a premium to traditional utility multiples. Bears counter that the $10 billion equity issuance planned for 2027-2030 will dilute existing holders just as borrowing costs reset higher in regulatory recoveries.

“Duke offers the cleanest pure-play exposure to the AI power buildout in the regulated universe,” said Nicholas Campanella, utility analyst at Barclays, in a research note circulated after the February call. “But the equity overhang is real. Investors should expect periodic pressure on the stock during issuance windows.” His view aligns with consensus across the sell side, where most analysts hold neutral-to-buy ratings with price targets clustered around $135 to $145.

The Data Center Surge Driving the Plan

The fundamental driver of the entire plan is a single brute-force trend: U.S. electricity demand growth has accelerated from a roughly flat decade between 2010 and 2020 to forecasts of 2-3% annual growth through 2030, with data centers responsible for roughly half of all incremental demand in 2025. That swing represents the most dramatic change in the U.S. power system since the postwar industrialization era.

Duke has disclosed that it received a 73% year-over-year increase in data center power requests in early 2026 alone, with individual project sizes climbing into the gigawatt range. The company expects retail sales growth of 1.5% to 2% in 2026, with that figure rising sharply as data center connections come online in 2027 and 2028. By 2030, Duke believes data center load could account for 25% or more of total system demand in its Carolinas footprint.

That trajectory has been validated independently by the U.S. Energy Information Administration and the International Energy Agency, both of which have raised their U.S. demand forecasts in successive reports. EIA’s Annual Energy Outlook now projects total U.S. electricity demand reaching roughly 5,500 TWh by 2030, and the IEA’s Electricity 2025 report confirmed that AI data centers are now the fastest-growing segment of global electricity consumption.

Regulatory Risk: Carolinas IRP and the NC WARN Pushback

Duke’s North Carolina utility filed an updated Integrated Resource Plan in late 2025 that explicitly underwrites the $103 billion expansion, projecting massive electricity growth driven by data center connections. The IRP is now under review at the North Carolina Utilities Commission, with formal hearings scheduled into the second half of 2026.

👁 Regulatory Risk: Carolinas IRP and the NC WARN Pushback

The April 21, 2026 NC WARN filing represents the most aggressive opposition Duke has faced in a decade. The group argues Duke is using ratepayer dollars to recruit data centers via discounted economic-development tariffs, and that the resulting generation buildout amounts to a “subsidized industrial policy” funded by households. Duke disputes the framing, pointing out that the company’s economic-development tariffs require minimum bill commitments and that data center loads ultimately spread fixed costs over a larger volume of sales.

Indiana presents a different but equally important regulatory front. Duke’s Indiana utility has filed for permission to begin construction on roughly 1.5 GW of new combined-cycle gas capacity, citing data center demand around the Indianapolis-Plainfield corridor. The Indiana Utility Regulatory Commission’s decision, expected in mid-to-late 2026, will set precedent for how Midwestern states allocate the cost of AI-driven generation.

Funding the Plan: $10 Billion in New Equity and Rate Base Growth

Funding a $103 billion plan requires more than retained earnings and incremental debt. Duke has explicitly told investors it intends to issue $10 billion in new equity between 2027 and 2030, layered on top of operating cash flow and a steadily expanding debt program. The equity ramp is sized to keep Duke’s regulatory equity ratio in line with the cost-of-capital structures approved by state commissions, currently anchored around 51% equity in the Carolinas.

The rate-base implications are profound. Total rate base across Duke’s regulated subsidiaries is projected to grow from roughly $84 billion at year-end 2025 to approximately $120 billion by 2030, an expansion that mechanically drives the company’s earnings growth. Every dollar of rate base earns a regulated return on equity – typically in the 9.5% to 10.5% range – meaning Duke’s earnings power scales almost linearly with successful capex deployment.

“This is the cleanest growth algorithm in U.S. utilities,” said Steve Fleishman, senior analyst at Wolfe Research, in commentary published shortly after the February announcement. “The plan does not require any merchant exposure, any unregulated bets, or any non-recurring asset sales. It is just rate-base growth funded by a balanced mix of debt, equity, and operating cash flow.”

Expert Analysis: What Industry Insiders Are Saying

Reaction across the utility analyst community has been broadly positive but laced with cautionary notes about execution risk and regulatory headwinds. The consensus view is that Duke has correctly diagnosed the demand opportunity but that the plan’s success hinges on three external variables: regulatory approvals, supply chain availability for gas turbines, and political acceptance of the resulting customer bill impact.

“Duke is essentially betting the company on the AI infrastructure thesis,” said Andrew Weisel, equity analyst at Scotia Howard Weil. “The upside is enormous if data center demand materializes as forecast. The downside is that if the AI buildout slows even modestly, Duke will be left with a generation portfolio sized for a load curve that never arrives.” Weisel cited the 2010s natural gas overbuild as a cautionary precedent, though he acknowledged that today’s data center contracts include far stronger commercial protections.

Maheep Mandloi, lead utilities analyst at Mizuho, took a more bullish stance. “The combination of secular load growth, regulated returns, and a defensible Southeast footprint makes Duke the single best risk-adjusted way to play the U.S. power transition,” he wrote in a March 2026 note. “Even if the plan slips by a year or two, the underlying earnings algorithm remains intact.” Mandloi has a buy rating on DUK with a $145 price target.

Comparing Duke’s Bet to Hyperscaler Power Strategies

Duke’s plan does not exist in a vacuum. The hyperscalers themselves have been racing to lock down power supply through a mix of utility partnerships, behind-the-meter generation, and direct power purchase agreements. Amazon Web Services has pursued nuclear-adjacent campuses through its Talen Energy deal in Pennsylvania and a separate effort with X-energy on small modular reactors. Microsoft signed a 20-year power purchase agreement with Constellation Energy to restart Three Mile Island Unit 1, expected online in 2028. Google entered a multi-site SMR partnership with Kairos Power.

👁 Comparing Duke's Bet to Hyperscaler Power Strategies

What distinguishes Duke’s bet from those hyperscaler-led initiatives is the regulated rate-base framework. The hyperscalers are funding generation through commercial contracts with merchant developers; Duke is funding generation through cost-of-service ratemaking, which spreads risk across millions of customers but also generates a regulated equity return on every dollar deployed. The hyperscaler model is faster but riskier; Duke’s model is slower but more durable.

The two approaches are also converging. Duke has begun negotiating “data center service agreements” that combine traditional retail tariffs with bilateral capacity reservations, minimum bill commitments, and pass-through fuel cost provisions. These hybrid contracts give hyperscalers tariff certainty while protecting other ratepayers from stranded asset risk – a structural innovation that may become the industry template by 2027.

2026-2030 Predictions: Where Duke Goes From Here

Looking ahead, the next five years will determine whether Duke’s $103 billion bet defines a generation of utility outperformance or becomes a cautionary tale of overbuild. The following predictions reflect the most likely trajectories based on current data.

• Prediction 1: Duke will raise its capex plan again at the February 2027 earnings call, likely past $115 billion, as additional data center connection requests crystallize into binding contracts.
• Prediction 2: The Belews Creek SMR site will be down-selected to a single reactor vendor by mid-2027, with construction permits filed in 2028 and commercial operation targeted between 2031 and 2033.
• Prediction 3: The Carolinas will adopt a formal data center tariff framework before the end of 2026, with minimum bill commitments and capacity reservation fees becoming standard for any load above 100 MW.
• Prediction 4: Duke will execute its $10 billion equity issuance program through a combination of forward sale agreements, at-the-market programs, and at least one block transaction, with the bulk completed by mid-2029.
• Prediction 5: By 2030, data center load will represent 25% to 30% of total Duke system demand, fundamentally reshaping the company from a residential-and-commercial utility into a hybrid hyperscaler-supply business.

Frequently Asked Questions

What is Duke Energy’s $103 billion plan?

Duke Energy’s $103 billion plan is the company’s 2026-2030 capital expenditure program, the largest fully regulated capital plan ever announced by a U.S. utility. It funds 14 GW of new generation, 4.5 GW of battery storage, transmission and distribution upgrades, and early development of a small modular reactor at Belews Creek in North Carolina.

When did Duke Energy announce the plan?

The $103 billion plan was first disclosed on the company’s February 11, 2026 fourth-quarter earnings call and reaffirmed in a Fortune feature published April 25, 2026. CEO Harry Sideris has positioned the plan as the largest fully regulated capital program in the industry.

How does Duke’s plan compare to NextEra and Dominion?

Duke’s $103 billion five-year plan exceeds NextEra’s roughly $75 billion plan, Southern Company’s $63 billion plan, and Dominion Energy’s $50 billion plan. On a per-customer basis, Duke is also the most aggressive spender among large U.S. investor-owned utilities.

What does Duke Energy’s stock cost in April 2026?

Duke Energy (NYSE: DUK) closed at $128.04 on April 23, 2026, giving the company a market capitalization of approximately $101 billion. The stock has traded in a 52-week range of $111.22 to $134.49.

Will customer bills go up because of the plan?

Customer bills will rise as Duke recovers its capital investments through state regulatory rate cases, but the company has emphasized affordability as a central pillar. Critics including NC WARN argue the plan adds roughly $60 billion to the Carolinas rate base over four years, while Duke counters that data center load growth ultimately spreads fixed costs over a larger sales base.

Is Duke Energy building a small modular reactor?

Duke Energy filed an early site permit application in December 2025 for a potential small modular reactor at the Belews Creek Steam Station in Stokes County, North Carolina. A reactor vendor has not yet been selected, but the site is positioned to host an SMR by the early 2030s.

How much new generation is in the plan?

The plan adds approximately 14 GW of incremental generation through 2030, including 5 GW of new natural gas combined-cycle plants in the Carolinas and Indiana, roughly 4 GW of utility-scale solar, and 4.5 GW of battery storage by 2031.

Who is Duke Energy’s CEO?

Harry Sideris is the CEO of Duke Energy, having succeeded Lynn Good in 2025. Sideris has framed the $103 billion plan as the largest fully regulated capital program in the U.S. utility industry and has emphasized the balance between data center growth and ratepayer affordability.

Why are data centers driving the plan?

U.S. data centers, fueled by AI workloads, drove approximately half of all U.S. electricity demand growth in 2025. Duke received a 73% year-over-year increase in data center power requests in early 2026, with individual project sizes climbing into the gigawatt range. The plan is sized to meet that surge.

How will Duke fund the $103 billion?

Duke plans to fund the program through operating cash flow, an expanding debt program, and approximately $10 billion in new equity issuance between 2027 and 2030. The funding mix is calibrated to maintain regulatory equity ratios required by state utility commissions.

What are analysts saying about the plan?

Analyst reaction has been broadly positive, with most price targets clustered between $135 and $145. Bulls cite Duke’s 9.6% earnings growth target and clean rate-base algorithm, while bears point to equity issuance overhang and political risk in state-level rate cases.

How does the plan affect U.S. cloud and AI infrastructure?

Duke’s plan is one of the largest single contributions to the U.S. utility power buildout supporting AI and cloud infrastructure. By underwriting 14 GW of generation and 4.5 GW of storage in data center hubs across the Carolinas, Indiana, and Florida, Duke is positioning itself as a critical enabler of hyperscaler expansion through 2030.

Related Coverage

• The $1.4 Trillion Grid Overhaul: 51 Utilities and $300B in AI Capex
• Half of U.S. AI Data Centers Delayed or Canceled: The 7 GW Capacity Crisis
• Big Tech’s $650 Billion AI Infrastructure Bet
• X-energy’s $1B IPO: Amazon-Backed SMR Bet for AI Data Centers
• Amazon’s $11.57B Globalstar Acquisition
• Cloud Computing 2026: The Pillar Guide

Sources: Utility Dive – Duke $103B, EIA Annual Energy Outlook, IEA Electricity 2025, NERC Reliability Assessments, Edison Electric Institute.