By; Robert Dugan, Director, Placer County Water Agency
On data readiness, supply chain transformation, and the governance advantage districts already have
Special districts exist because communities decided that certain infrastructure responsibilities are too important, too technical, and too long-term to be managed well by general government. Water. Sanitation. Fire protection. Recreation. Resource conservation. The governance model is built on the premise that focused accountability produces better outcomes than diffuse responsibility.
Artificial intelligence is about to test that premise in a new way — and I think special districts are better positioned to meet that test than most people in this sector yet realize.
I write from two chairs simultaneously. As a director of the Placer County Water Agency, I sit on a board that governs public infrastructure for one of California's fastest-growing regions. As CEO of CalCIMA, I represent the producers of aggregate, ready-mix concrete, asphalt, and industrial minerals that build the pipelines, treatment plants, roads, and facilities that special districts operate. I see AI arriving from both directions at once — as a tool for governance and operations, and as a force reshaping the supply chains districts depend on for their capital programs.
What is actually arriving — and what isn't.
The noise around AI has created two equally wrong impressions: that AI is going to transform everything immediately, and that it's mostly hype that doesn't apply to districts like ours. The reality is more specific and more useful than either.
AI tools that are already shipped and in use across infrastructure organizations include: predictive maintenance platforms that identify equipment failure signatures weeks before a failure occurs; AI-assisted permitting and document review that cuts processing time significantly; demand forecasting models that incorporate climate variability and land use change in ways no spreadsheet can replicate; and real-time procurement intelligence that for the first time is creating price discovery in materials markets that have historically run on relationships and phone calls.
What is not yet here for most special districts: the "agentic AI" that executes multi-step administrative workflows autonomously. That is coming — it is already shipping in large construction management software platforms — but governing boards and general managers can take a breath. The more urgent question is not when the sophisticated tools arrive. It is whether your district's data will be ready when they do.
The data tax your district is already paying.
Jim Manning, Chief Customer Officer at CDE Lightband — a municipal public power utility — put a name to something special districts know by feel even if they haven't named it. He calls it the "data tax": the administrative cost paid every day by organizations whose operational information lives in disconnected silos. Systems that don't talk to each other. Reports that require manual assembly. Decisions made on last month's numbers because real-time data isn't accessible.
His utility avoided $1,250,000 in demand charges in a single fiscal year using AI-driven load forecasting. But his opening argument wasn't about the AI tool. It was about what he called data hygiene and centralization — the discipline of connecting systems, cleaning information, and eliminating the silo structure that makes the data tax so expensive. The AI worked because the data foundation was ready.
For special districts, the data tax shows up in capital planning gaps, in reactive maintenance rather than predictive maintenance, in procurement cycles based on historical relationships rather than real-time market intelligence. These are not technology problems. They are data discipline problems — and they are solvable with the systems many districts already have, if those systems are connected and maintained.
Garbage in, Armageddon out. The stakes in our industries are physical and regulatory, not just operational. Getting the data right isn't optional. It's the foundation everything else stands on.
What AI is doing to the supply chains your capital program depends on.
Here is the dimension of AI that I hear almost no discussion of in special district circles, and it matters directly for capital planning and procurement.
The materials industries that supply special district capital programs — aggregate, ready-mix concrete, asphalt — are themselves in the early stages of AI transformation. Meta and Amrize — formerly Holcim's North American business — released an open-source AI model this year that produced concrete reaching full structural strength 43% faster with 35% lower embodied carbon, tested at scale on a real construction project. AI-driven mix design is not a research concept. It is shipping today.
Separately, AI-powered materials sourcing platforms — Bulk Exchange being the leading example in aggregates and bulk construction materials — are beginning to create real-time price discovery in a market that has historically had none. The implications for capital project cost estimation are significant. If your district's estimating process is based on last cycle's supplier relationships rather than current market intelligence, the gap between your budget and your bid results is going to widen.
Neither of these changes requires special districts to do anything immediately. But they do require district staff and governing boards to ask a new question when reviewing capital program budgets: are we pricing this project against the market that exists, or the market we knew six months ago?
Govern it like you govern everything else that matters to your community.
Colorado Springs Utilities CEO Travas Deal presented at a recent utility industry conference what may be the most practically useful AI governance framework I've encountered for public agencies. He calls it Crawl, Walk, Run.
Crawl: identify low-risk use cases, run small pilots, define your guardrails before you need them. Walk: controlled deployment with real governance, clear ownership, and accountability. Run: AI managed as a genuine organizational portfolio, with the same rigor applied to any other enterprise-level investment.
His information classification system is simple enough for any district to adopt immediately: Public information is fine for any AI tool. Internal Use Only information is limited to approved, governed tools. Confidential information — rate data, personnel records, legal strategy, infrastructure vulnerabilities — never goes into an AI tool. Period. Three categories. An afternoon of staff discussion. Done.
|
The governance parallel that matters most for special districts:
We don't build safety plans to prevent work from getting done. We build them so work can get done at full speed, safely, without catastrophic failure. The same logic applies to AI governance. The districts that will move fastest and furthest with these tools are the ones that establish their guardrails early — before a bad outcome forces the issue. Innovation and discipline are not opposites. In our sector, they never have been.
|
Why special districts are better positioned than they think.
The special district governance model — focused accountability, direct community connection, long-term stewardship horizon — is not a liability in the AI transition. It is an asset.
Large private companies adopt AI fast but govern it poorly. General government moves slowly and often misses the operational moment entirely. Special districts occupy a middle position that should allow them to move deliberately, govern rigorously, and adopt in a way that serves their communities rather than their quarterly earnings.
That positioning is most valuable when it is used proactively. The districts that will shape how AI serves public infrastructure in California are the ones that start the internal conversation now — about data readiness, about vendor governance, about procurement intelligence, about what questions to ask before the next capital program budget is submitted.
AI arrived in our world without an invitation. The adaptation question is not whether to respond. It is whether to respond on terms we set, or terms set by vendors and technology waves we didn't design and weren't consulted about.
Special districts were created because focused accountability produces better outcomes. That principle applies to AI adoption exactly as it applies to everything else we do.
Whether a district is just beginning to explore AI or already developing a governance framework, the conversation belongs in your boardroom — not just in your IT department. According to a recent Procore survey, 48% of construction leaders — the people building and maintaining the infrastructure special districts own and operate — identify AI as the most critical technology for future project delivery. That assumption is already embedded in the construction interface your capital programs depend on.
CSDA's governance resources, leadership programs, and peer network exist precisely for moments like this one. The districts that will shape how AI serves California's communities are the ones that start that conversation before the technology makes the decision for them.
Robert Dugan
Director and former Chair, Placer County Water Agency
President, Sierra College Foundation
President & CEO, California Construction and Industrial Materials Association (CalCIMA)
Robert Dugan brings more than four decades of experience at the intersection of California public policy, water governance, infrastructure, and the industries that build and sustain the state's built environment. He has served in the California Legislature, led public affairs for one of California's largest construction companies, and served as a civic and educational leader across the Sacramento/Placer County region.
#FeatureNews