Whitepaper
AI, Sovereignty, and the Information War
A Fact-Based Look at Foreign Influence, Data Center Economics, and the AI Policy Choices Ahead
Dan Bond · Founder & Principal AI Advisor, LumenForge Advisors LLC · September 2026
A note on method
Every claim in this document is labeled by confidence tier — Confirmed (multiple independent, attributable sources), Verified with caveats (documented but contested, single-sourced, or regionally variable), or Contested/Evolving (proposed policy not yet enacted, or actively disputed). Sources with a financial or advocacy stake in the outcome are disclosed where relevant.
Executive Summary
Bottom line: Three things are true at once. Treating them as a single story is where most coverage of this topic goes wrong.
First, there is a documented, multi-vector effort to shape American opinion against U.S. AI infrastructure — a covert social-media operation OpenAI disrupted in 2026, and a separate track under Senate Intelligence Committee referral involving state media and foreign-tied advocacy funding. Those two tracks are related in theme. They are not the same evidence base, and this paper does not treat them as one.
Second, those campaigns did not invent American skepticism of data centers. They tried to ride concern that already existed and runs deep — including in York County, where a QTS campus, a nine-month moratorium, and neighbors organizing around noise, water, and rates are a real, homegrown fight with nothing to do with foreign influence.
Third, the risk calculus cuts both ways. Nation-state-developed AI models, particularly from China, carry documented security, censorship, and data-jurisdiction risks. Any U.S. policy response that unilaterally slows American AI development without a credible mechanism to bind competitors does not reduce that risk — it reallocates it.
This document lays out the evidence for each claim, tiered by confidence, so it can inform decisions rather than reinforce a conclusion reached in advance in either direction. The full whitepaper covers the two influence tracks, the national data on data-center electricity costs, nation-state AI model risk, the compliance-asymmetry problem in current policy proposals, and what this means for decision-makers.
Dan Bond is Founder and Principal AI Advisor at LumenForge Advisors LLC. He spent 26 years at Microsoft, most recently as Director of AI Strategy & Transformation for the Americas, and now advises small and mid-sized organizations in the Carolinas on making AI usable, governable, and theirs.
This paper is an educational briefing, not legal, engineering, or investment advice, and not a statement on any pending land-use application.