A Kentucky Town Tried AI for Civic Engagement — And It May Have Changed Everything
Earlier this year, Warren County, Kentucky hosted a month-long town hall unlike any other — and nearly 8,000 residents participated. The secret behind the record-breaking turnout? Artificial intelligence.
Bowling Green, the third-largest city in Kentucky and part of Warren County, is bracing for a major population boom by 2050. Local officials knew they needed public input to plan for the future, but traditional town halls weren’t cutting it. These meetings often attract only a handful of residents — typically those with strong, sometimes negative, views — leaving most voices unheard.
Online surveys could theoretically solve this, but they generate massive amounts of data that are difficult for humans alone to sort through and analyze meaningfully. That’s where AI came in.
Bowling Green’s Digital Town Hall
To gather input at scale, county leaders turned to technology. In February, they partnered with a local strategy firm to launch an online platform where residents could anonymously share their ideas for the city’s future. The tool used was Pol.is, an open-source survey platform known for its success in civic engagement, particularly in Taiwan.
The prompt was simple: What would you like to see in your community over the next 25 years? Residents could submit suggestions and vote on others’ ideas. Over 33 days, around 8,000 people participated, submitting over 4,000 unique suggestions and casting more than a million votes.
Ideas ranged from new museums and expanded pedestrian walkways to more green spaces and restaurants. The participation level — roughly 10% of Bowling Green’s population — was unprecedented for a civic exercise.
How AI Made Sense of It All
The collected data was fed into Sensemaker, an AI tool developed by Google’s tech incubator Jigsaw. Sensemaker categorizes public input, identifies key themes, and evaluates consensus across responses. From thousands of submissions, the AI highlighted 2,370 ideas that had at least 80% support.
Some of the most widely supported proposals included:
- Increasing access to healthcare specialists to reduce reliance on facilities in nearby Nashville
- Repurposing empty retail spaces for community use
- Adding more dining options to the city’s north side
The insights were compiled into a public policy report, now available on the project website. “If I have a town hall meeting on these topics, 23 people show up,” said Doug Gorman, Warren County Judge Executive. “And what we just conducted was the largest town hall in America.”
Reaching the Unreachable
The digital approach allowed officials to connect with demographics that usually go unheard — including working-class residents, political independents, and immigrant communities. The survey supported multiple languages and automatically translated responses, making it accessible to non-native English speakers.
One local advocate, Daniel Tarnagda, an immigrant from Burkina Faso and nonprofit founder, praised the inclusive format. “I knew that people want to be part of something. But if you don’t ask, you don’t know,” he told PBS.
Now, volunteers are distilling the report into actionable recommendations, expected to be delivered to county leadership by year’s end. According to a Jigsaw-conducted survey, the use of AI tools saved local officials an average of 28 work days compared to manual processing.
Beyond Bowling Green: Agreement Without Politics
This marked the first large-scale test of Sensemaker in the U.S., and Jigsaw described it as a successful proof of concept. One of the most revealing findings? When ideas were presented anonymously, stripped of political affiliations, residents discovered they agreed on far more than expected.
“When only the loudest or angriest voices are heard, we develop a caricatured view of the other side,” said Yasmin Green, CEO of Jigsaw. “One of the most important things AI can do is help us keep the conversation going.”
Building on this momentum, Jigsaw announced a partnership with the Napolitan Institute, founded by pollster Scott Rasmussen. The new project aims to survey Americans across all congressional districts — not for policy, but to understand shared national values and hopes for the future.
The Risks of AI in Democracy
While the Bowling Green project has been widely praised, experts warn of potential pitfalls in using AI for civic engagement. Despite clear disclaimers that no personal or demographic data was collected in this initiative, concerns remain about privacy and data security in future implementations.
If responses were ever tied to personal identities or leaked, participants could face harassment or discrimination based on their political beliefs. There’s also the issue of algorithmic bias — where the views or biases of developers may unintentionally shape AI output. A recent example: Elon Musk’s AI chatbot, Grok, was found to reflect Musk’s personal views in its responses to sensitive topics.
Such biases could make AI-generated policy suggestions less objective — and, in some cases, deeply flawed.
A Glimpse Into AI’s Democratic Potential
Despite these concerns, Bowling Green’s experiment demonstrates that AI can dramatically expand civic participation, particularly among historically underrepresented groups. And when used transparently and ethically, it may even help bridge divides in polarized communities.
The initiative didn’t just ask people what they wanted — it created space for agreement, collaboration, and shared vision. In an era of fractured politics and low trust, that might be AI’s most promising contribution yet.
