Editor: Anayana White, ASBN Head of Communications
As artificial intelligence (AI) sweeps across the globe, promising to revolutionize everything from medicine to daily productivity, a massive physical infrastructure is quietly being constructed behind the scenes. This digital expansion requires a staggering amount of electricity, and it raises a critical, often ignored question: Who is paying for the massive grid upgrades needed to support this AI boom?
To answer this question, the Pennsylvania Sustainable Business Network (PASBN), in partnership with the American Sustainable Business Network (ASBN) and Compound Impact Collective, has officially released a landmark new report, Energy Affordability for Pennsylvania’s Small Businesses: Data Centers & Energy Choices in the AI Era. This report, funded by the Heinz Endowments, is a crucial first step in examining the direct and indirect impacts of rapid AI data center development on small and medium-sized businesses in Pennsylvania as a case study to understand the impacts that are sweeping across the country.
A recent webinar hosted by ASBN brought together experts to discuss the report findings. The panel featured David Levine, President and Co-founder of ASBN; Michael Jarrett, lead report author and former Deputy Director of PASBN; Jorge Luis Fontanez, lead analyst and Founder of Compound Impact Collective; and was moderated by Michael Green, ASBN Senior Policy Advisor.
Watch the launch webinar
Key Takeaways from Who Pays for the AI Boom for Business Leaders
Here are the key takeaways, notable statistics, and policy recommendations from the newly released report and webinar discussion.
A Voice for Main Street in the AI Era
For too long, conversations surrounding AI and data centers have focused strictly on tech-sector growth or residential consumer protections. Our new report addresses the critical missing piece: how the AI boom impacts local small businesses and ratepayers.
With over 1.1 million small businesses in Pennsylvania employing 2.5 million individuals (over 50% of the state’s private workforce), safeguarding these enterprises from escalating energy costs is not just a local issue; it is vital to the entire Commonwealth’s economic stability.
Small Businesses are Subsidizing Tech Giants
One of the report’s most striking findings is that everyday ratepayers, including residents and local small businesses, are currently carrying the financial burden of preparing the electrical grid for tech giants. Jorge Luis Fontanez explained:
“One of the headlines here is that small businesses are subsidizing the infrastructure of some of the most valuable companies in the world.”
In 2024, Pennsylvania ratepayers absorbed roughly $492 million in data center infrastructure costs embedded in their utility bills. That pattern isn’t unique to Pennsylvania. Across the PJM region that same year, 124 of 130 data center transmission projects—about 95%—were paid for by ratepayers rather than by the data centers that triggered them, according to the Union of Concerned Scientists. The companies building these facilities are largely not covering the grid upgrades their demand requires, while local utility customers are.
Why Small Businesses are Uniquely Vulnerable to AI Data Centers Energy Rate Hikes
While large corporations have the resources and institutional power to navigate energy markets, local small businesses do not.
Large tech companies utilize dedicated procurement teams to negotiate long-term Power Purchase Agreements (PPAs) and secure lower, preferred energy rates due to the sheer volume of electricity they consume. Small businesses, however, are lumped in with general ratepayers and must buy electricity at standard market rates.
This leaves small businesses highly exposed to price volatility. As Michael Jarrett explained:
“Historically speaking, small businesses operate with razor-thin margins, so even a 10% increase in annual energy costs can have a significant impact on their ability to hire staff and employ more people in the state, grow and expand their businesses, or in some cases, just stay afloat.”
Despite their massive economic importance, small businesses have largely been left out of energy affordability conversations because they lack the collective voice to negotiate with utilities. PASBN, ASBN and Compound Impact Collective designed this report to serve as that collective voice.
AI Data Center Disproportionate Impacts on Communities
The physical footprint of the AI boom is also creating geographic and environmental tensions:
- Loss of Agricultural Land: Land is being sold to data center developers at high markups, raising serious concerns about food security.
- The “Race-to-the-Bottom” Paradigm: Wealthier-income communities often leverage their power to block data centers. Consequently, these massive projects get pushed into disadvantaged or lower-income communities that often lack resources, connections to media and legislators, and the time to organize around a collective voice, exacerbating existing environmental injustices.
- Strained Water and Watersheds: Data centers require millions of gallons of water to cool their equipment, threatening local watersheds decades into the future.
Policy Interventions: Driving Transparency and Accountability in the AI Era
How do we balance the economic potential of the AI sector without harming local communities and small businesses?
The report outlines several regulatory mechanisms and models:
The “Bring-Your-Own-Energy” (BYOE) Model
Instead of hooking up directly to the public grid and draining existing power resources, PASBN advocates for policies that require data centers to build or procure their own localized, clean energy sources (such as colocating with solar farms or installing rooftop solar arrays).
Community Benefit Agreements (CBAs)
The authors pointed to Lancaster, Pennsylvania as a model of community engagement. There, a data center developer signed a CBA that established two funds totaling $20 million—one for local economic development and the other a sustainability fund to invest in solar infrastructure on municipal buildings.
State Legislation on the Table
The report highlights a patchwork of evolving state-level policies across the nation, emphasizing that Pennsylvania must act:
- Pennsylvania House Bills:
- Other State Precedents:
- Ohio: Approved a tariff requiring data centers above 25 MW to pay for at least 85% of their capacity for up to 12 years.
- New Jersey: Enacted the Data Center Fair Share Act to ensure large facilities contribute equitably.
- Oregon: Passed the Oregon POWER Act of 2025, enacting large-load tariffs.
- Virginia: Imposed a per-kilowatt-hour tax on data center electricity to generate public tax revenue.
The Urgency of the AI Data Center Clean Energy Transition
Currently, Pennsylvania has no clean energy procurement requirements for data centers, in stark contrast to states like New York (which mandates 70% clean energy) and Michigan (90%).
To close this gap, PASBN and its partners are backing the proposed PRESS (Pennsylvania Reliable Energy Sustainability Standard) and PACER (Pennsylvania Climate Emission Reduction) programs. If adopted, these bills would require 35% of utility energy to come from clean sources by 2035, while returning 70% of revenues to ratepayers as rebates. This program is projected to save ratepayers an estimated $252 million and create nearly 15,000 jobs.
Conclusion: In the AI Era, Small Businesses & Communities Get a Seat at the Table
The decisions made about energy and infrastructure in this decade will dictate community structures for the next 30 to 40 years. To avoid a system where small businesses and communities are excluded, policymakers must act swiftly.
As David Levine summarized:
“We believe this report is a first step [in] understanding the impacts on small and medium-sized businesses, and we offer that these businesses should have a seat at the table.”
Download the Report & Get Involved
- Download the full report: https://asbnetwork.org/pasbn/#Report
- Tell Your Story: Are you a small business owner experiencing rising utility costs? Are you concerned about a newly proposed project or existing data center in your area? Participate in ASBN’s “Tell Your Story” campaign to share the impacts you are facing.
The opportunities are endless! The AI boom, including the advances in AI tools and the infrastructure needed to support this sector, opens the door for entrepreneurs to find innovative solutions that we don’t even know exist yet. This could be anything from more efficient, less expensive battery storage to more efficient servers requiring less space and energy.
It depends on how “cost” is defined, but yes. The development costs for BYO (Bring Your Own) energy would vary. Energy Innovation, a non-partisan energy and climate policy think tank, concluded that a combination of clean energy sources and on-site energy storage is the fastest, cheapest, and most reliable way to power AI data centers.
Energy Innovation, a non-partisan energy and climate policy think tank, concluded that a combination of clean energy sources and on-site energy storage is the fastest, cheapest, and most reliable way to power AI data centers. If data center developers use this approach and invest in clean energy and storage at the projected scale needed for AI data centers, both will become even more affordable.
Gov. Shapiro’s executive order requires local officials to sign off on proposed data centers. As a result, these projects will likely be discussed in public hearings where local citizens and businesses can comment on the proposal. This process will likely vary from place to place depending on local government structure. It has the potential to create much more transparency in the decision-making process, however, we don’t know the particular details yet for how local officials will have authority to deny a project. It may be safe to assume that a public comment period is linked to this, but the question of local authority remains and is outstanding. We encourage businesses and communities to reach out to PA legislators with this in mind.
The research did not explicitly distinguish between small businesses and minority-owned small businesses; however, the report does note that traditionally disadvantaged communities, which can include higher percentages of minority-owned businesses, are disproportionately impacted. Future research, including surveys of the small business sector, could yield results that help better answer this question.
Water use related to data centers is a potential problem; however, the water use of every data center in America in 2023 equaled less than 1% of the water used to irrigate the nation's agriculture that year, according to reporting by Axios. The real concern is watersheds in the areas where data centers are being developed, including both stress on water sources as well as pollution. This can and should be part of the community benefit conversations. Alternatively, legislation to hold data center developers to higher standards when it comes to their impact on local water supplies. A secondary concern report authors see is how water supplies would be affected by expanding the use of fossil fuels to power data centers. And third, while we don’t cover it in the report, the use of PFAS in data center cooling processes is of growing concern. Gov. Shapiro’s executive order does require data center developers to adhere to high water-conservation standards, but does not address the use of PFAS or other toxic chemicals.
The fast-tracking of data centers was definitely being experienced earlier this year; however, the politics around this have changed. As a result, development plans are undergoing additional scrutiny, including standard environmental assessments for these types of developments. Gov. Shapiro’s new executive order requires that water and energy disclosures be filed before a project is approved as well as states that data centers have been removed from the commonwealth’s fast-track development programs, and developers can no longer use nondisclosure agreements (NDAs), which can be used to help developers skirt assessments, development regulations, and keep communities in the dark. However, the EO does not account for noise, air pollution or other quality of life measures which might be relevant to an environmental assessment.
###
About American Sustainable Business Network
American Sustainable Business Network (ASBN) amplifies the collective voice of sustainable business to lead the way to a regenerative economy that is stakeholder-driven, just, and prosperous. As a multi-issue, membership organization advocating on behalf of every business sector, size, and geography, ASBN works to advance its mission to inform, connect, and mobilize sustainable business leaders, transforming the public and private sectors toward a just and regenerative economy.