Ethical Dilemmas in Balancing The Pros & Cons of AI Data Center Locations

Meeting Overview

Download Data Center Ethics Presentation (PDF)

These notes summarize the November 11, 2026 discussion hosted by the HHI Ethics Society (Hilton Head Island) regarding the ethical dilemmas, infrastructure strains, and regulatory challenges of building AI data centers. The group weighs the broader benefits of AI against local environmental, economic, and governance impacts—particularly in South Carolina and the Lowcountry applying the utilitarian principle that actions are right if and only if they create the greatest amount of good coupled with the least amount of harm.

Key Themes & Discussion Points

Scale & Types of Data Centers

  • Unlike traditional small/medium data centers, hyperscale data centers built for AI training chain thousands of processors together, multiplying power and cooling demands by 10x to 100x.

  • Neoclouds rent space in existing facilities to lease AI capacity to third parties, though some face financial uncertainty (such as CoreWeave) compared to massive hyperscalers like Amazon, Microsoft, Google, Oracle, and Meta.

  • Many AI applications require near real time communications that require data centers to be geographically distributed to serve the user populations rather than placed arbitrarily far away.

Benefits of AI vs. Ethical & Technical Risks

  • Benefits: Advances in healthcare diagnostics, environmental conservation and wildlife tracking (such as bird-song monitoring), accessibility/education, disaster response, and data analysis.

  • Risks & Behaviors: AI models can "cheat" or deceive by bending guardrails to fulfill tasks fast, cybersecurity risks/AI "escapes" from testing environments, privacy concerns with personal/commercial AI agents, and "garbage in, garbage out" errors (including high-stakes military/drone contexts).

Environmental & Infrastructure Strain

  • Massive demands on local water (for cooling) and electricity (including proposals tied to coal power expansion or long-timeline nuclear options), plus a looming "tsunami" of electronic waste as hardware becomes obsolete rapidly.

  • Siting data centers in space or underwater faces severe heat-dissipation, maintenance, and repair challenges.

  • Local communities—especially smaller ones with fewer resources—often lack leverage or clear oversight when large projects are proposed, prompting moratoriums (noted in places like Spartanburg, Anderson, and elsewhere in South Carolina, as well as a stopped proposal near the ACE Basin).

Regulation, Incentives & Community Trade-offs

  • Participants debated local zoning/land-use power versus calls from major tech companies for one nationalized regulation, raising concerns about regulatory capture by a few large companies with strong lobbying/campaign influence, and whether these companies might shift liability onto taxpayers.

  • While companies may offer taxes, infrastructure, roads, or community/school contributions, members noted permanent staffing at large facilities can be small compared to temporary construction jobs, and promised water/power/noise mitigations may still strain regional resources.

Historical Context & Next Steps

  • The group compared AI to past technological shifts (electricity, PCs, the internet, social media), noting that while the underlying pattern of innovation vs. disruption isn't new, the speed and exponential scale of problems make oversight urgent.

  • The next meeting is scheduled for the first Wednesday in February, tentatively focused on elections and the proposed “National Public Vote Law” for President and Vice President.

 

Next
Next

Separation of Church and State: Democracy, if you can keep it