By Better Together Agency
International Monetary Fund Managing Director Kristalina Georgieva put the moment plainly on October 12: “AI, love it, hate it. It’s already here.”
I understand why this line gets attention. AI is part of work, research and daily decisions. Companies are investing heavily in the technology behind it. But the fact that AI is here does not settle who benefits, who pays or what communities get to decide.
Georgieva was discussing the role of AI infrastructure investment in global economic growth, including potential gains in Asian economies. Those forecasts deserve attention. So do the people living near proposed facilities, whose questions cannot be answered by a global growth forecast.
The International Energy Agency reported in April 2026 that capital spending by five major technology companies exceeded $400 billion in 2025. The same report noted that some proposed data center projects will never be built.
Large investment plans do not settle the merits of each site. Electricity supply, local infrastructure, water availability and the terms offered to residents differ from one location to another. A project with public support in one place may face legitimate objections in another.
The IEA also cautions against a simple claim that data centers always increase electricity prices. The outcome depends on the local power system, new infrastructure requirements and policy decisions. That is precisely why communities need information about the project proposed for them.

A Gallup poll published in May 2026 found 71% of U.S. adults opposed constructing an AI data center in their area. In the same March poll, 53% opposed constructing a nuclear energy plant in their area. That’s an 18 percentage point difference. Nearly half, 48%, strongly opposed the AI data center.
I find that comparison striking. Data centers are often described as necessary infrastructure for AI growth. Yet more people opposed a data center near their home than a nuclear plant. Gallup asked parallel questions about local construction, and the results show how much resistance AI infrastructure faces. They do not tell us why each respondent took that position.
Gallup also found 70% worried a great deal or a fair amount about the environmental impact of AI data centers. In a separate April survey, opponents cited electricity and water use, pollution, utility bills and quality of life. Supporters pointed to jobs and other potential economic benefits.
A person may use AI at work and still question a proposed facility near their home. A resident may support new business activity while disagreeing with a project’s water requirements or financial terms.

I would want to see answers to five questions.
First, what electricity and water resources would the facility need, and how were those estimates prepared? Plans should account for changes as the facility grows.
Second, who would pay for new power lines, substations, roads and other improvements? Communities need to understand what could appear on utility bills or public budgets.
Third, what local economic benefits can be documented? Construction jobs, long term positions, taxes and incentives should be reported separately. Forecasts should be identified as forecasts.
Fourth, what information will remain public after a facility opens? Reporting should address resource use, compliance and commitments made during approval.
Fifth, when can residents influence the decision? Public meetings held after the terms are settled leave little room for participation. Companies and public officials should explain what alternatives were considered and what could still change.
Disclosure is a starting point. Trust depends on how companies answer questions, handle disagreement and follow through.

These questions do not end at a data center fence. Organizations using AI should know which tasks it can assist with, which materials it may access and where a person must review its work. They should be able to explain how they check facts and potential bias before decisions affect others.
My work on AI bias and governance has reinforced a basic principle. People must have enough information and authority to question a system, correct it or decline to use it for a particular purpose.
The same principle applies to infrastructure decisions. Public participation is part of accountability. It should not be treated as a barrier to progress.
AI is already here. Communities still deserve a say in what comes next.