*This article, written by Butler Snow attorneys Tray Hairston and Charity Karanja, was originally published on The Bond Buyer.
AI data centers are often framed as a technology story.
Increasingly, however, they are a public finance story.
The rapid expansion of artificial intelligence infrastructure will
influence where new tax base is created, how utilities plan for
unprecedented load growth, how communities manage increasingly
constrained power and water resources, and how the municipal
market finances the infrastructure that connects private investment
to public benefit.
Those themes were at the center of a June 9, 2026, panel at the
National Association of Securities Professionals’ 37th Annual
Financial Services Conference in Detroit, Michigan. The session,
“The AI Data Center Supercycle: Capital Markets, Incentives,
Power & Water: How Wall Street and Main Street Can Win
Together,” brought together leaders from public power, investment
banking, economic development and local government to examine
the opportunities and challenges emerging from the next wave of
infrastructure investment.
Moderated by Dr. Dell Gines, senior advisor and former Chief Innovation Officer at the International Economic Development
Council, the panel featured Adam Barsky, Executive Vice
President and Chief Financial Officer of the New York Power
Authority (NYPA); Harry Singh, Managing Director at Goldman
Sachs; Neal Richardson, Director of Housing and Economic
Development Advisory Consulting at Baker Tilly; and Joseph
Deason, Executive Director of the Madison County Economic
Development Authority in Mississippi.
Rather than focusing on the magnitude of the market’s investment
in AI, the discussion centered on the practical realities shaping the
sector’s future, including infrastructure constraints, community
readiness and the financing mechanisms required to support
sustained development.
Power as the Defining Constraint
Like the infrastructure revolutions that preceded it, the AI data
center buildout is poised to reshape America’s and the world’s
economic landscape. Railroads, electrification, interstate highways and telecommunications networks each redirected capital,
commerce and opportunity which often produces uneven outcomes
across regions and communities.
The AI data center buildout differs in one critical respect: access to
capital is not the primary constraint. It is the industry’s access to
power that is causing that logjam. Singh illustrated the scale of the
opportunity by noting that while the telecommunications boom of
the late 1990s generated roughly $300 billion in capital formation,
AI-related data center investment could ultimately approach $5
trillion. U.S. installed data center capacity, currently estimated at
approximately 40 gigawatts, is projected to more than double by
the end of the decade.
These are not incremental investments. They represent a
fundamental expansion of the nation’s digital infrastructure. As a
result, the primary question for the site selection industry has
become increasingly straightforward: can a community provide
reliable power at the scale and speed hyperscale operators require?
Barsky described the growing mismatch between development
timelines and electric grid realities. A hyperscale facility requiring
hundreds of megawatts of capacity, often representing billions of
dollars in investment, cannot always wait years for traditional
interconnection processes and transmission upgrades to be
completed.
That reality is reshaping energy strategies across the sector.
Behind-the-meter generation, fuel cells, natural gas-fired
generation and phased power solutions are increasingly being
deployed to bridge the gap between immediate demand and
long-term grid expansion. While effective in the near term, these
approaches also raise important questions regarding cost allocation,
reliability, emissions and long-term system planning.
Over the longer term, small modular nuclear reactors may offer a
more durable answer. Their scale is well suited to hyperscale data
center demand, and they provide the reliable, emissions-free
baseload generation that increasingly constrained grids will
require. Yet commercial deployment remains years away, making
them unlikely to alleviate the immediate power shortages
influencing today’s investment decisions.
The broader lesson is that energy infrastructure has become
economic infrastructure. Communities that can deliver reliable,
affordable power will be best positioned to compete for AI-related
investment, while those that cannot may find themselves
increasingly sidelined.
Energy infrastructure has become economic infrastructure.
For issuers, utilities and public finance professionals, power is no longer a secondary consideration. It is becoming one of the
primary determinants of growth, credit strength and long-term
economic opportunity.
Traditional Economic Development Metrics Are Under Strain
Data centers challenge many of the assumptions that have long
shaped economic development policy.
Unlike manufacturing facilities, distribution centers or corporate
headquarters, hyperscale data centers are extraordinarily
capital-intensive but comparatively light on permanent
employment. They require significant investments in land, power,
water and transportation infrastructure, yet the number of
long-term jobs they generate may appear modest relative to the
scale of the investment. As a result, traditional measures of
economic impact do not always tell the full story.
Richardson noted that communities are frequently asked to provide
zoning approvals, tax incentives, utility commitments and political
support for projects whose benefits may not be immediately visible
to residents. Evaluating these investments solely through the lens
of direct job creation can leave local officials struggling to explain
the value proposition to taxpayers.
Deason offered a perspective informed by Madison County,
Mississippi, where hyperscale investment has emerged as a
significant economic development catalyst. Over the past two and a
half years, the county has announced more than $21 billion in
private investment commitments and approximately 1,700 jobs.
While those jobs are important, Deason emphasized that the
broader fiscal impact may ultimately be more transformative. The
expansion of the county’s tax base is expected to support
investments in schools, transportation, water and wastewater
infrastructure, broadband expansion and other public assets that
strengthen both quality of life and long-term economic
competitiveness.
That distinction is particularly important for policymakers and
public finance professionals. The value of a data center investment
cannot be assessed solely through direct employment figures.
Rather, it should be evaluated through its broader economic and
fiscal contribution including tax base growth, infrastructure
improvements, utility system enhancements, workforce
development initiatives, local procurement opportunities and the
long-term financial capacity it creates for local government.
In many respects, the rise of data centers challenges communities
to rethink how economic development success is defined. For
projects of this scale, the more meaningful measure may not be the
number of jobs created within the facility itself, but the extent to
which the investment expands public resources, strengthens critical infrastructure and creates enduring economic value for the
community as a whole.
Water Is Emerging as a Critical Site Selection Factor
While power remains the primary constraint on data center
development, water is rapidly becoming an equally important
consideration.
As communities increasingly scrutinize the environmental and
infrastructure impacts of hyperscale facilities, water consumption
has moved to the forefront of the site selection process and local
permitting discussions. Cooling technologies, conservation
practices and long-term water sourcing strategies are now central
components of the development process.
Deason pointed to Madison County, Mississippi, as an example of
how communities are beginning to address water constraints
through infrastructure innovation. Several hyperscale operators are
utilizing reclaimed water sourced from treated residential and
industrial wastewater rather than relying exclusively on potable
water supplies. Through advanced treatment processes, including
ultrafiltration and reverse osmosis, reclaimed water can be purified
to a standard suitable for data center cooling operations.
This distinction is quite important. As water availability becomes a
more prominent public concern, communities are increasingly
scrutinizing projects that place additional demands on drinking
water systems. In the alternative, investments that expand
reclaimed water capacity, modernize wastewater treatment
infrastructure and improve overall system resilience can align
economic development objectives with responsible resource
management.
For bond market participants, water is no longer simply an
operational consideration. It is increasingly a factor in project
feasibility, permitting risk, infrastructure investment requirements
and long-term credit analysis. As competition for data center
investment intensifies, communities with sustainable water
strategies and the infrastructure to support them may enjoy a
meaningful advantage in attracting and retaining large-scale
development.
The Capital Stack Extends Well Beyond the Hyperscaler
Data center development is often viewed through the lens of large
technology companies deploying capital directly into new
facilities. In reality, the market has evolved into a far more
sophisticated ecosystem involving merchant developers,
infrastructure investors and a diverse range of capital providers.
Increasingly, facilities are developed on a speculative or pre-leased basis with hyperscale operators serving as long-term customers
under service-level agreements. These agreements typically impose
rigorous performance and availability standards supported by
meaningful financial remedies in the event of service disruptions.
As a result, they create highly predictable contractual revenue
streams that can underpin a variety of financing structures.
The predictability of those cash flows has important implications
for capital formation. Long-term service agreements can support
project finance structures as well as attract institutional capital and,
in certain cases, facilitate asset-backed securitizations and other
capital markets transactions. As the sector matures, data center
financing is becoming increasingly sophisticated which broadens
the pool of capital available to support growth.
Barsky highlighted another financing tool with direct relevance to
the municipal market: prepaid power transactions. NYPA recently
completed nearly $1 billion of prepaid solar transactions in which
tax-exempt bonds were used to prepay for power generation. By
leveraging the differential between taxable and tax-exempt
borrowing costs, the structure reduced the effective cost of energy
to the end user by an estimated 7% to 10%. One recent transaction
involved Google as the power purchaser.
The significance of these transactions extends beyond any single
project. They illustrate how municipal finance tools can help align
public infrastructure assets with private-sector demand, creating
opportunities to reduce costs, improve efficiency and support
long-term investment.
The same principle applies to the broader infrastructure required to
support data center growth, including transmission upgrades,
substations, water and wastewater facilities, transportation
improvements and utility extensions. Where hyperscale
development creates durable sources of tax or utility revenue, local
governments may be better positioned to finance infrastructure
investments that would otherwise be difficult to undertake.
Deason noted that Madison County is preparing a significant
public infrastructure financing supported, in part, by the expanded
tax base generated through hyperscale investment. The example
highlights a broader trend: as AI infrastructure investment
accelerates, the municipal market is likely to play an increasingly
important role in financing the public assets that enable
private-sector growth.
For issuers, advisors and investors, understanding the evolving
capital stack behind data center development may prove just as
important as understanding the facilities themselves. The most
significant opportunities may lie not only within the projects, but in
the infrastructure systems that make those projects possible.
Community Benefit Cannot Be an Afterthought
If power is the defining technical challenge of the data center
buildout, public trust may be its defining political challenge.
Across the country, proposed data center developments have
encountered moratoriums, litigation and growing community
opposition. In many cases, that resistance is not rooted in
opposition to growth or investment. Rather, it reflects uncertainty
about who will benefit, who will bear the costs and whether the
promised economic returns will ultimately materialize.
Richardson argued that the challenge is often structural rather than
ideological. Too frequently, communities are asked to evaluate
complex projects only after key decisions have already been made.
Local officials are then left to explain complicated questions
involving power demand, water consumption, tax incentives and
land use under intense public scrutiny and compressed timelines. A
more effective approach begins well before a developer arrives.
Communities that establish clear frameworks for evaluating major
projects are often better positioned to negotiate outcomes that align
with local priorities. Those frameworks may address workforce
development, local contracting opportunities, infrastructure
investments, utility cost allocation, revenue sharing, environmental
performance, transparency measures and accountability provisions.
Such preparation benefits both sides of the transaction. It provides
elected officials with a clear basis for evaluating and explaining
proposed investments, while offering developers greater certainty
regarding community expectations and approval processes.
Viewed through that lens, community benefit agreements and
similar public benefit structures are not simply negotiating tools.
They are mechanisms for managing risk. By establishing
expectations early, they can reduce delays, strengthen public
confidence and improve the long-term durability of a project.
Deason emphasized the importance of translating large-scale
investments into outcomes that residents can readily understand. In
Madison County, community support grew as residents began to
see the connection between hyperscale investment and tangible
public benefits.
Discussions about billions of dollars in capital investment became
more meaningful when framed in terms of expanded school
funding, improved roads, enhanced water infrastructure and
broader access to broadband.
The lesson is straightforward. Community engagement is not a
procedural step to be completed after a deal has been structured. It
is a fundamental component of project success. Communities are
far more likely to support transformative investments when they
understand how those investments advance local priorities and when they are treated as participants in the process rather than
observers of it.
Community engagement is not a procedural step. It is a fundamental component of project success.
What the Municipal Market Should Do Now
The AI data center buildout is advancing at a pace that often
outstrips traditional public-sector planning and decision-making
processes. The appropriate response is not to accelerate approvals
at the expense of diligence. Rather, it is to begin planning earlier
and engage more strategically.
For public finance attorneys, municipal advisors, investment banks,
issuers and economic development professionals that starts with
viewing data centers as infrastructure transactions as much as
economic development transactions. The most important questions
are often not about incentives alone, but about the public
infrastructure required to support long-term growth.
Can the electric grid support projected demand? How will water
resources be managed? What infrastructure investments will be
necessary to accommodate development? Who will bear those
costs? What revenues will support public financing? How will
incentives be measured and enforced? And how will local
governments protect themselves if investment projections fall short
of expectations?
The panel discussion highlighted the broader significance of these
investments. Barsky noted that data centers are increasingly
viewed not simply as economic development projects, but as
critical infrastructure with implications for national
competitiveness and security. The communities that host the
physical infrastructure supporting artificial intelligence will play an
important role in the next phase of economic growth. Securing that
opportunity, however, requires more than attracting private capital.
It requires a deliberate strategy for translating investment into
lasting public benefit.
The moderator Gines and panelists also challenged the perception
that only major metropolitan areas are positioned to compete. In
many cases, rural and secondary markets may possess advantages
that larger markets increasingly lack including available land,
lower development costs, greater flexibility in infrastructure
planning and stronger community alignment. Madison County’s
experience illustrates that there are several communities that are
willing to do things to understand the industry and its needs in
addition to preparing proactively and negotiating from a position of
strength and clarity. This type of approach can successfully attract transformational investment.
For the municipal market, the opportunity extends beyond
incentives. Prepaid power structures, utility revenue bonds, tax
increment financing, infrastructure financings supported by
expanding tax bases and other public finance tools can help
communities leverage private-sector investment to build public
assets that create value long after a project is completed, for
example new schools, parks, libraries, better roads and other
infrastructure improvements.
The AI data center expansion is not a future possibility. It is
already underway. Communities across the country will experience
its effects either directly or indirectly. The more important question
is whether local governments, utilities, economic developers and
capital markets participants will work together to shape those
outcomes or whether communities will find themselves reacting
after key decisions have already been made.
Detroit provided a fitting setting for that discussion. Few places
understand more clearly both the transformative potential of
infrastructure investment and the challenges of aligning that
economic growth with broad-based community benefit.
The next major infrastructure cycle has already begun. The
municipal market has an opportunity, and a responsibility to help
ensure that that cycle is structured in a way that creates lasting
value for communities and investors alike.
