Policy Pressures the AI Power Trade Toward Execution Phase
ENERGY POLICY PERSPECTIVES: VOL. 22
The political and regulatory environment around AI data centers is shifting the AI power trade from broad thematic exposure toward execution. Political and regulatory scrutiny is increasing and, as demonstrated by the recent actions of Pennsylvania Governor Josh Shapiro and Texas Governor Greg Abbott, policymakers are adding more conditions on AI data center development as they seek clearer answers on project credibility, cost responsibility, electric reliability, and local benefits. This raises execution risk and reduces the value of ambitious AI load growth projections and potential AI data center pipelines, but does not necessarily weaken the underlying AI power thesis, in our opinion. Instead, we expect value to accrue to companies able to convert AI load growth into contracted, physically deliverable, and politically resilient projects while penalizing those whose AI exposure proves largely aspirational. This favors utilities that can convert load into politically palatable infrastructure growth as well as IPPs that can solve site-specific power constraints.
IPPs: Fewer speculative GWs, but more value per credible GW. Greater political scrutiny may reduce realized AI load, delay timing, and raise execution risk, but data centers that survive may place greater value on electricity physically deliverable at their location, by the required COD with the required reliability. This favors IPPs able to monetize scarce existing generation, brownfield optionality and interconnection access, as well as broader physical power-solution capabilities to work around constraints to help enable speed-to-power.
Key risks amid rising political pressure: 1) lower realized load if marginal projects fail; 2) potentially delayed project timelines; 3) political intervention in electricity markets; and 4) greater execution burden to provide packaged power solutions.
Key mitigants: 1) secure well-structured agreements with creditworthy large loads; 2) scarce existing generation, brownfield sites, uprates and assets located near constrained load; 3) diversified market exposure; 4) greenfield development capabilities and secured equipment; and 5) ability to provide integrated capabilities.
Utilities: Special contracts may be the more politically durable solution. For utilities, the opportunity remains substantial, but the risk-sharing model is changing as existing ratepayers do not want to be held liable for the AI infrastructure buildout. Some states have already established large-load tariffs, contractual protections and transparent processes to manage AI development, while others are only now developing frameworks, creating a wide variance in political risk exposure among utilities. Rising political concerns around the socialization of AI-specific infrastructure and risks increase the need for cost allocation frameworks that clearly protect residential ratepayers as well as the ability to demonstrate net-benefits to other rate classes, states and local communities.
Key risks amid rising political pressure: 1) less ability to socialize incremental costs; 2) greater scrutiny of large-load forecasts and capital plans; 3) waning political popularity of AI influences capital deployment; and 4) customer protections could limit the risk/reward versus traditional rate base treatment.
Key mitigants: 1) special contracts and large load tariffs; 2) phased and/or customer-backed investment; and 3) minimum bills, collateral and termination protection; 4) constructive state regulatory and political environment; and 5) ability to coordinate and structure bundled solutions for faster energization.
AI data centers are the physical, visible manifestation of AI. We believe AI data centers, as well as the electrical infrastructure they require, have become the physical, visible embodiment of AI, an otherwise largely invisible technology. In our view, this helps explain why broader anxiety around AI is increasingly showing up in large load policy. The underlying asymmetry should be familiar to AI investors, the costs of the AI buildout are immediate and local while many of its benefits remain prospective, diffuse, and uncertain. The emerging response is not necessarily to stop AI and data centers (NY is the only state to successfully pass a statewide moratorium to date), but to change the political bargain to require developers to internalize more of the infrastructure, reliability and local costs they create as well as providing tangible benefits in return for a “social license” to exist. Hyperscalers appear willing to pay premiums for speed-to-power, redundancy, infrastructure and a “social license” today, but that willingness to pay may not be unlimited.
Political durability is becoming part of the investment case. Rising local opposition to data centers is giving politicians across the aisle greater incentive to revisit the terms of AI development as illustrated by Governor Shapiro’s August 18th order and Governor Abbott’s August 3rd audit. The grievances are converging, but the state regulatory mechanisms are still far from equivalent, elevating execution risk as obligations vary by jurisdiction and can evolve over the life of a multi-decade asset (see Figure 1). Pennsylvania is tightening permitting and development conditions, while Texas has added a verification screen to ERCOT’s newly established large-load process. For hyperscalers and developers, the key for site selection may be to not wait for permanent clarity that will likely never come, but favor jurisdictions where the project can remain economically and politically defensible even as political leadership and economic circumstances change.
Figure 1: Converging Grievances, Different Regulatory Mechanisms
From socialized to contracted reliability. The most important implication for utilities and IPPs is the shift on who bears which risks. Political pressure is pointing toward utilities becoming more contractual and IPPs becoming more infrastructural (see Figure 2). Clearer allocation of risk can improve the bankability of bilateral structures even as regulation becomes more demanding. This does not eliminate political risk, but we think it makes projects more defensible by clearly assigning costs, protecting incumbent customers and reducing strandedasset exposure. Policy is also pushing credible loads to identify and underwrite their physical power solutions or accept uncertain service. In our opinion, this increases the value of scarce generation, deliverable capacity and power-rich sites while pushing utilities, IPPs and large loads toward more long duration power arrangements.
Figure 2: Old vs New Utility and IPP Paradigms
Alignment is a core challenge. The political challenge at its core is one of alignment, not purely on a partisan basis, but institutional as well as jurisdictional (see Figure 3). The Trump administration places greater weight on national security, US AI leadership and economic competitiveness, while governors, state utility commissions and local governments directly face rising electricity bills, zoning disputes, water use and constituent opposition. The structural tension is that federal policy increasingly seeks to internalize AI’s costs so development can accelerate, while state and local policy seem more willing to accept slower or altered AI deployment to protect their constituents. Meanwhile, utilities and IPPs want to build while reaching internal return hurdles and managing downside risk, hyperscalers want to move as fast as possible (potentially with or without the grid) but are constrained by states and ISO/RTOs are increasingly stretched trying to manage this fragmented system.
Figure 3: Players, Incentives and Frictions Shaping AI Data Center Development
Near term catalysts convert politics into market design. The policy debate is increasingly flowing through two distinct but overlapping channels: ISO/RTO rules governing whether and how large loads can reliably connect, and state/utility frameworks governing who pays and bears the risk of serving them. While there are numerous proceedings ongoing nationwide, our near-term catalyst watch is focused on PJM and ERCOT, where large-load growth is driving particularly fluid and market-sensitive decisions around resource adequacy, interconnection and service eligibility (see Figure 4). At the state level, the starting points vary widely. Some have already established large load frameworks for accommodating AI data centers in a politically and economically viable manner, while others are only now defining how costs, service rights and risks should be allocated. These proceedings produce fewer headlines, but can materially affect capital deployment, customer economics and project bankability.
Figure 4: Catalyst Calendar
Analyst Certification
We, Christopher R. Ellinghaus and Gabriele Sorbara hereby certify that the views expressed in this research report accurately reflect our personal views about the subject companies and their securities. We further certify that no part of our compensation was, is, or will be directly, or indirectly, related to the specific recommendations or views contained in this research report.
Financial Interests: Neither we, Christopher R. Ellinghaus or Gabriele Sorbara, nor any member of our households own securities in any of the subject companies mentioned in this research report. Neither we, nor a member of our households is an officer, director, or advisory board member of the subject company or has another significant affiliation with the subject company. We do not know or have reason to know at the time of this publication of any other material conflict of interest.
Analyst Compensation: The authors compensation is based upon the value attributed to research services by Siebert Williams Shank institutional brokerage clients. The authors of this report are compensated based on the performance of the firm, and have not received any compensation in the past 12 months from any of the subject companies mentioned in this report. The performance of the firm is driven by its secondary trading revenues, investment banking revenues, and asset management revenues.
SIEBERT WILLIAMS SHANK EQUITY RESEARCH RATINGS KEY
BUY: In the analyst's opinion, the stock will outperform the S&P 500 on a total return basis over the next 12 months.
HOLD: In the analyst's opinion, the stock will perform in line with the S&P 500 on a total return basis over the next 12 months.
SELL: In the analyst's opinion, the stock will underperform the S&P 500 on a total return basis over the next 12 months.
Distribution of Equity Research Ratings as of: August 21, 2026
IBC (Investment Banking Clients) is defined as companies in respect of which Siebert Williams Shank (the “firm”) or its affiliates have received or are entitled to receive compensation for investment banking services in connection with transactions that were publicly announced in the past 12 months.
__________________________________________________________________________________________________________________
Other Important Disclosures
Investment Banking Disclosures: Within the past 12 months, the research analyst authoring this report has not participated in a solicitation of any subject company mentioned within this report, with or at the request of investment bankers, for investment banking business. Within the past 12 months, the firm and its affiliates have not managed or co-managed a public offering of the securities of any subject company mentioned within this report, nor has the firm received compensation for investment banking products or services from those companies.
Firm Compensation: Within the past 12 months, the firm and its affiliates have not received compensation for any non-investment banking products or services for subject companies mentioned in this report, and none of these subject companies have been a client of the firm during the past 12 months.
Stock Ownership: The firm and its affiliates do not own 1% or more of any class of equity security in this report, and do not make a market in any such securities.
Disclaimers: The information and opinions contained in this report were prepared by the firm and have been derived from sources believed to be reliable, but no representation or warranty, expressed or implied, can be made as to their accuracy. All opinions expressed herein are subject to change without notice. This report is for information purposes only and should not be construed as an offer to buy or sell any securities. The firm makes every effort to use valuation methodologies that it believes to be reasonable in the derivation of price targets, but we do not guarantee that such methodologies are accurate.
__________________________________________________________________________________________________________________
Additional Information: To receive any additional information upon which this report is based, please contact: Siebert Williams Shank, Research Department
100 Wall Street, 18th Floor
New York, NY 10015
212-830-4500
