TL;DR

  • AI is already embedded in hyperscale data center program management — synthesizing portfolio data, generating risk briefings, and driving capital decisions across portfolios of 50 or more active construction sites simultaneously.
  • Unlike AI in financial services, healthcare, or aviation, data center program management AI operates with no external governance framework, no auditability standards, and no defined accountability structures.
  • The industry needs an AI governance framework built on four pillars: auditability of AI-generated decisions, risk-based oversight design, data governance for sensitive infrastructure intelligence, and clear accountability structures.
  • Organizations like the Uptime Institute and the Project Management Institute are well positioned to lead this work — and the industry does not need to wait for regulatory intervention to begin.

# # #

The data center industry is in the middle of the most consequential infrastructure build cycle in modern history. Hundreds of billions of dollars in capital investment. Hundreds of active construction sites across every major market simultaneously. Power grids strained. Permitting systems overwhelmed. Labor markets stretched thin.

And increasingly, the program operations managing all of this are being run — at least in part — by AI.

Large language models now synthesize portfolio data, generate executive briefings, flag risk conditions in construction timelines, and accelerate decision-making for the technical program managers who orchestrate these massive programs. This is not a future scenario. It is the present reality of how hyperscale infrastructure is being built and operated today.

What is notably absent from this reality is any governance framework designed specifically for it.

The Governance Gap Nobody Is Talking About

When AI is deployed in financial services, it faces regulatory scrutiny. When it is deployed in healthcare, it faces auditability requirements. When it is deployed in aviation, it faces rigorous certification standards. When AI is deployed to manage a portfolio of data center programs — facilities that will house the compute powering global financial systems, healthcare records, and logistics networks for the next two decades — it operates with essentially no external governance requirement whatsoever.

This is not a theoretical concern. It is a live operational reality with downstream consequences that the industry has not yet seriously engaged with. Consider three specific governance failures already materializing across hyperscale programs:

  • AI-generated recommendations flow into capital decisions without documented reasoning or defined accountability for outcomes.
  • Automation bias goes unmanaged in high-pressure, high-volume environments where program managers handle 30, 50, or more active sites simultaneously.
  • Sensitive infrastructure intelligence — facility locations, power capacity commitments, security posture data — is ingested by AI systems governed only by internal policy, with no industry-wide standards for access control or data retention.

A Framework Built on Four Pillars

The data center sector has mature frameworks for physical security, power redundancy, environmental compliance, and construction safety. It needs equivalent standards for the AI systems now embedded in program operations. Four pillars provide the foundation:

  • Auditability — Every AI-generated decision in a critical program context should be logged, reviewable, and traceable to source data. AI systems must produce decision records as a standard output, not as an afterthought.
  • Oversight Design — Human review requirements should be defined by risk class. A schedule optimization warrants different oversight than an AI-generated risk assessment informing a site acquisition decision. Glancing at a briefing and approving it is not oversight.
  • Data Governance — Sensitive infrastructure data ingested by AI platforms — facility locations, power commitments, security posture — should be subject to explicit classification, access controls, and retention standards. Organizations should require vendors to answer: who has access, and is this data used to train the model?
  • Accountability Structures — Clear chains of accountability for AI-influenced outcomes should be documented as part of standard program governance. When a flawed AI recommendation drives a capital write-down or a delayed program, the organization needs to know who was responsible and what they knew.

How the Industry Can Move Forward

The Uptime Institute, the Project Management Institute, and infrastructure industry associations already set operational standards for data center programs. They are the natural home for this work — and they do not need to wait for regulatory intervention to begin. A working group focused on minimum requirements for AI auditability, oversight, data governance, and accountability would give the industry a foundation it can implement immediately.

The facilities being built today will define the digital economy for decades. The AI managing those programs deserves governance frameworks commensurate with that responsibility.

Conclusion

Hyperscale data center AI is a present operational reality, not a future consideration. The governance vacuum is real, consequential, and addressable without waiting for regulators or high-profile failures. The industry needs a working group, a baseline framework, and the will to treat AI governance as seriously as it treats power redundancy. The infrastructure is too important for anything less.

# # #

About the Author

Sanchita Hosur is a Technical Program & Product Leader with 12+ years of experience in global data center infrastructure, AI-powered program operations, and enterprise technology delivery. She has led AI platform programs serving 200+ global stakeholders across hyperscale infrastructure portfolios spanning AMER, EMEA, and APJC. Her research focuses on AI governance frameworks for critical infrastructure program management. She holds an MS in Technology Management from the University of Illinois Urbana-Champaign and a BE in Electrical & Electronics Engineering from Visvesvaraya Technological University.