Michigan is becoming a live case study in artificial intelligence’s physical footprint.
Since the launch of ChatGPT in November 2022, ethical conversations about AI have focused on how people use the tools. Should AI be used for content creation? How much information should someone share with AI tools? When should someone disclose AI usage? How do we prevent AI from spreading misinformation or reinforcing biases? Those questions deeply matter.
They matter because AI is already shaping how people learn, work and think critically. Workers are asking if AI will replace jobs or reshape the skills needed to keep them. Educators are asking what AI means for learning and future readiness. Organizations are asking how to adopt AI to improve efficiency, streamline operations and stay competitive. Meanwhile, the public conversation often swings between hype and fear, making opportunities and risks harder to separate.
So, who am I to be talking about this? Fair question.
I’m no one’s engineer and shouldn’t be asked to build a server. But I spent much of my undergraduate and graduate career at Michigan State University studying the ethical and social implications of generative AI in public relations under the mentorship of Dr. Chuqing Dong, assistant professor of Advertising and Public Relations at MSU.
Over three years of qualitative research, we examined how generative AI is reshaping the public relations industry through professionals’ own experiences, including new opportunities and ethical considerations around trust, disclosure, accountability and workplace well-being. Our work earned the Top Ethics Paper Award at the 2024 International Public Relations Research Conference and was later published in Public Relations Review.
I carried those ideas into my master's thesis, which explored dialogue theory and how workplaces can create the conditions for more open, inclusive and ethical conversations about generative AI. I have also presented and served as a panelist at regional, national and international events, including BledCom in Slovenia, MSU Ignite Talk, PRSA chapter programs and to the MSU Board of Trustees.
Now back to what I was saying.
AI has the potential to expand accessibility, productivity gains and knowledge-sharing as it becomes increasingly embedded in daily life. But those benefits don’t erase uncertainties, and workforce concerns show why. While some research shows U.S. entry-level job postings have fallen, largely due to AI, describing AI as a major factor behind recent layoffs is often more complex than headlines suggest. Companies may cite AI as a reason for workforce changes while also responding to broader economic pressures or shifting business priorities. In my own research, professional communicators voiced similar concerns that AI may serve as a convenient explanation for workforce reductions, even when reasons for those decisions are more layered.
Although AI may not be the sole cause of every workforce change, it’s becoming a visible symbol of disruption. AI adoption is moving at sprint speed as policies and training still find their footing. So it’s no wonder nearly three-quarters of Americans say AI is moving too fast. Public concern about AI usage and its economic gains cannot be dismissed as politics or resistance. If anything, it signals a widening trust gap between implementation and accountability.
As AI is sold for efficiency and speed, communities are asking about the hard cost: What does “efficiency” require?
That question moves the ethical conversation beyond tool use toward the inner workings of the systems. How does AI work? Where do its answers come from? What does it take to power them? Who benefits, and who bears the impact?
Now, these concerns are showing up in Michigan communities through data centers.
Data centers power the AI systems people use every day. Google AI search summaries, chatbot conversations and AI-generated images depend on servers that need land, electricity, cooling and facilities to run — and lots of it. In other words, every prompt has a physical reality behind it. The convenience we experience on-screen depends on infrastructure that has to exist somewhere, and that “somewhere” is increasingly up for public debate.
This is where the next ethical communications challenge begins.
Saline Township shows the tension between investment and public trust
The Saline Township data center, known as The Barn, has become the nucleus of Michigan’s AI debate.
The project is massive. Developed by Related Digital for Oracle and its customer OpenAI, it’s expected to support the AI systems more people and businesses use every day, including tools like ChatGPT, as well as AI-powered search, customer service, software development and data analysis platforms.
State and business leaders frame it as a defining economic opportunity for Michigan, with thousands of union construction jobs, hundreds of permanent onsite jobs, tax revenue for schools and the township and new investment in local services. Gov. Gretchen Whitmer called the $16 billion project “the largest one-time investment in state history.”
To supporters, The Barn positions Michigan as a serious competitor in the next generation of technology, innovation and economic development.
For those opposed, the data center raises concerns about water and energy use, air quality, child health and electricity bills. When agricultural land turns into a data center site, opponents also worry about losing the ground that supports fresh food and working livelihoods.
Saline Township officials voted 4-1 against rezoning agricultural land for the development. The township later settled after Related Digital and landowners sued, allowing the project to move forward with conditions. For a small town of less than 3,000 people, continuing the fight meant weighing public opposition against the high legal costs and risks of taking on a much larger developer. The settlement included funding for farmland preservation, area fire services, water protections, limits on noise and future expansion, community investment and a fund set aside to help cover the costs if the facility shuts down.
Those protections are significant, but the process that produced them deserves scrutiny.
When a community’s elected representatives vote against a project and the project advances anyway, the communications challenge evolves. Public concerns were heard, perhaps, but not powerful enough to change the outcome. In a rural community like Saline Township, that feeling can deepen mistrust.
No one will argue Michigan shouldn’t compete for innovation and economic growth. However, communities and those who represent them as elected officials deserve a clear understanding of what that growth requires.
Data centers may bring investment, jobs and tax benefits. They may also create pressure on water systems, electric grids, farmland, ratepayers and local trust.
In the Great Lakes state, water carries history and identity. Only 1% of Great Lakes water is naturally renewed each year, which makes every major withdrawal part of a larger, interconnected system. When data centers rely on drinking water, lakes and rivers, developers should provide clarity on how resources will be used and protected.
Clarity is ever more important when a project’s impact is difficult to envision in a neighborhood. Some reports estimate a medium-sized data center alone can use about 100 million gallons of water per year for cooling servers, roughly the annual water use of 1,000 U.S. households. Some might point out other industries or habits that use more water, including fast fashion, where just one cotton shirt requires 700 gallons. But that misses the point. One high-volume use doesn’t justify another. It should push every major water user toward stronger conservation and accountability for shared water resources.
Ethical AI communication means clarifying implications and involving impacted groups in the conversation before decisions are advanced. Even better, it means developing responsible solutions to the real concerns and problems, and building them into the process from the beginning.
Michigan can support innovation and still ask hard questions. The true measure of progress isn’t how many data centers the state can attract. It’s whether Michigan can prosper in ways that are transparent, accountable and responsive to the communities most affected.
Allen Park shows what happens when context is missing
Allen Park offers a different view.
The city's planning commission denied a proposed data center site plan after postponing the vote three times. Residents raised concerns about environmental impact, energy demands, noise and public health.
The outcome didn't simply reflect opposition to the development. It reflected the lack of information provided to decision-makers and residents about the project’s impact.
That is the heart of the issue. Residents are being asked to evaluate a fast-moving technology with incomplete information, uneven expertise and high stakes. The public is also being asked to trust claims that are often technically complex, biasedly presented or hard to compare due to limited national examples and research.
Michigan communities are already slowing the process to understand what those impacts could mean locally. Right here near Lansing, Meridian Township became one of about 50 Michigan communities to place temporary limits on data centers, with pauses lasting between 180 days and one year. State lawmakers have also proposed suspending data center projects until spring 2027, and the Michigan for Responsible Data Centers coalition offers resources to help communities, local officials and planners make sense of data center proposals to inform decision-making.
Taken together, these efforts show how difficult the public conversation has become. Communities are constantly surrounded by information on the news and in their social feed, but not all information is, well, informative or easy to verify. Projected water and energy use is often hidden through nondisclosure agreements, folded into municipal utility reporting or difficult to separate from indirect water use tied to power generation. When people have too much noise and not enough explainable information, it’s harder to trust how a project will actually affect their own backyard.
Ultimately, data requires context before anyone can claim it as accurate or false. A number without a plain-language explanation isn't transparent if it doesn't create public understanding. A public meeting after community concern erupts doesn't replace proactive dialogue. And that gap has consequences.
People are often excluded from a conversation when they don't understand enough to know what questions to ask. Silence also doesn’t always signal consent. It can reflect confusion, or be a symptom of intimidation discouraging someone’s feeling of safety in raising their voice.
Productive disagreement belongs in the process
Communities are being asked to absorb huge infrastructure decisions before they fully understand the cumulative impact.
These decisions can affect the very places people live: water, utility bills, job opportunities, property values, farmland, traffic patterns, family well-being and sense of control over their community’s future.
This is where professional communicators have a responsibility that goes beyond messaging.
Consensus shouldn't be the only goal. In complex public decisions, disagreement can reveal new information left out of another view. Communities need space to surface overlooked risks, challenge assumptions, invite curiosity and bring lived experience into the decision-making process.
It also requires recognizing power. A small township and a multi-billion dollar developer don’t enter a legal or public process with the same resources. That imbalance determines who can keep fighting and who feels pressure to compromise even when there is public opposition.
Public input already shapes decisions people experience in daily routines. Communities weigh in on new bike lanes, park and recreation amenities, road projects and sound walls along highways as they drop their kid off to school or go to work. While those conversations aren’t always simple, disagreement is still expected. They’re familiar functions of government, with established expectations of how residents can weigh in and how leaders should respond.
On the other hand, data centers are immersed in uncertainty. The conversation is loud, shaped by AI hype, political tension, outside pressure and conflicting information. Residents can have the opportunity to speak, but speaking isn’t the same as being equipped to participate meaningfully. When relationships are already strained, unanswered questions feel heavier. Open, inclusive dialogue can help leaders make more informed decisions with public trust on the line.
If AI is becoming part of public life, then its infrastructure must be part of PRODUCTIVE public dialogue. After all, innovation and accountability are interconnected.
Professional communicators can help organizations listen before positions are solidified, translating complex information into societal implications. As liaisons, we can build stronger bridges between organizations and impacted communities so public input shapes decisions before conflict becomes the only way people are heard.
