Google is expanding its use of artificial intelligence into one of the most complicated areas of public-sector technology: helping local governments understand what their residents actually want.
Image Courtesy : google.com
Through Google.org, the company is supporting a new Jigsaw Partner Program designed to bring Jigsaw's Sensemaking AI tools to more local governments, allowing officials to conduct large-scale civic conversations and use artificial intelligence to organize and analyze the feedback they receive.
The initiative represents another step in Google's broader effort to demonstrate how AI can be applied to government services. Instead of using AI primarily to automate administrative work, Jigsaw's project focuses on the relationship between governments and the people they serve, giving communities new ways to provide input while giving officials tools to process much larger volumes of feedback.
Google announced the program on September 22, 2026, describing it as part of Google.org's commitment to helping governments use AI to improve public services. Google.org is joining philanthropic organizations including Bloomberg Philanthropies and the David & Lucile Packard Foundation to support the expansion of the program.
At the center of the initiative is Jigsaw, a Google incubator that develops technology aimed at addressing challenges involving information, security and human agency. Jigsaw describes one of the problems it is trying to address as a “deficit in structural agency”—the gap between people's desire to influence the systems around them and their practical ability to do so.
The new program is built around a relatively simple premise: governments often receive far more public feedback than their employees can realistically process manually.
A city may conduct a community survey, hold a public meeting or launch an online consultation and receive thousands of individual responses. Those responses can contain valuable information, but sorting through them can require substantial staff time.
Jigsaw's Sensemaking AI is designed to provide a layer of analysis between the public conversation and government officials.
The technology can identify, categorize and summarize topics, themes and areas of alignment within large collections of responses. Jigsaw's documentation says its sensemaking library uses Google's Gemini models and can process conversation data within minutes, giving organizations a way to extract themes and other insights from large-scale deliberations.
The goal is not for an AI model to make policy decisions.
Instead, the technology is intended to help humans understand what residents are saying.
That distinction is important because the new program arrives as governments are experimenting with increasingly capable AI systems across a wide range of functions. Google Public Sector has separately promoted AI for administrative workflows, data analysis and service delivery, while Google.org's $30 million AI for Government Innovation challenge is supporting 15 organizations developing AI applications for public-sector problems.
Jigsaw's approach addresses a different part of the government technology equation.
Rather than primarily helping officials process paperwork or automate internal tasks, Sensemaking AI is focused on helping officials process public input.
That makes the program particularly relevant to local governments, where decisions about growth, infrastructure, transportation, housing and community services can generate thousands of competing opinions.
One of Jigsaw's most prominent examples comes from Bowling Green, Kentucky.
The city used an AI-enabled town-hall process to engage residents as part of planning for the community's future growth. According to Google, approximately 8,000 residents participated, generating thousands of ideas and more than one million responses from other participants. Jigsaw's tools were then used to analyze the conversation and help local leaders understand the themes emerging from the public discussion.
The Bowling Green experiment illustrates why the technology could be attractive to cities.
A traditional public meeting is constrained by physical space and time. Even a large online survey can force residents to choose from questions created in advance by whoever designed the survey.
A deliberative conversation can work differently.
Residents can explain their views in their own words, respond to other people's ideas and raise issues that may not have been anticipated when officials initially designed the engagement process.
The resulting information can be much more complicated than a collection of yes-or-no answers.
That complexity is also where AI becomes useful.
A human team attempting to read thousands of individual comments might identify broad themes, but doing so consistently and quickly can be difficult. A machine-learning system can process the text at much greater scale and identify recurring topics or areas of agreement that might otherwise take considerably longer to uncover.
Jigsaw's research has been exploring this concept beyond individual cities.
The organization has also conducted “We the People,” a nationwide conversation involving more than 2,500 Americans. Participants were able to express their views in their own words, respond to other participants and vote on propositions generated from the broader conversations using AI. Jigsaw reported that more than 90% of participants felt their opinions were represented in the resulting summary, while nearly 80% said they gained a better understanding of other participants' perspectives.
The experiment reflects Jigsaw's broader argument that AI could potentially combine the scale of polling with some of the qualitative depth of focus groups.
Traditional polling can collect responses from large populations but often reduces complicated opinions to predefined questions and answers.
Focus groups can produce richer discussions but typically involve relatively small numbers of people.
AI-assisted civic conversations attempt to bridge those two models.
People can provide open-ended responses at large scale, while AI helps researchers and officials identify patterns across the resulting discussion.
That capability could become increasingly relevant as local governments try to involve residents in complex decisions.
Consider a city preparing a long-term growth plan.
Residents might agree that growth is inevitable while disagreeing about where new development should occur. Some might prioritize transportation infrastructure. Others might focus on housing affordability, schools, parks or environmental considerations.
A conventional survey could ask residents to rank those issues.
A large conversation could reveal something more nuanced: how residents connect those issues to one another and why particular priorities matter to them.
AI-assisted analysis could then help officials see those relationships.
Jigsaw's work in Bowling Green was specifically tied to long-term community planning. Google says the resulting input helped inform a 25-year growth plan for the city. Similar initiatives are now underway in Chattanooga, Tennessee; Riverside, California; Kitchener, Ontario; and other communities.
The new Partner Program is intended to make that kind of capability available to more local governments.
Google says participating localities will be able to use the AI tools to run large-scale civic conversations without operational costs through the program. The company is encouraging governors, mayors and other local officials around the world to express interest in participating.
That philanthropic funding model is significant.
Local governments frequently operate under tight technology budgets, making it difficult to experiment with emerging AI systems even when those systems could potentially improve public services.
Providing the technology without operational costs could lower one barrier to experimentation.
At the same time, governments adopting AI for civic engagement will have to consider how the technology affects the interpretation of public opinion.
AI-generated summaries are not raw public opinion. They are outputs produced through an analytical process.
That process can involve choices about which comments are grouped together, which themes receive attention and how disagreements are represented.
An AI system could potentially overlook a minority viewpoint, merge distinct positions into a single theme or give disproportionate prominence to topics that appear frequently.
For that reason, AI-assisted civic engagement does not eliminate the need for human review.
Instead, it changes the task.
Rather than asking government employees to manually read every response, the technology can provide an initial analytical layer that officials and community participants can examine and question.
Transparency could therefore become an important part of the system.
Residents may want to know how their responses were processed, what AI models were used, how summaries were generated and whether humans reviewed the results before they were incorporated into a planning process.
Jigsaw's decision to develop open-source tools could provide one avenue for greater transparency and experimentation.
The organization's broader work emphasizes sharing research, datasets and code and treating its projects as starting points rather than finished answers. Jigsaw says its objective is to build free tools that give people greater voice and choice in the systems affecting their lives.
The privacy implications are another consideration.
Large-scale civic conversations can contain personal experiences, identifying information and sensitive opinions. Governments and organizations using these tools will need to determine what information should be collected, how long it should be retained and who should have access to it.
The larger the conversation becomes, the more important those safeguards become.
There is also a broader question about representation.
An AI system can analyze the people who participate, but it cannot automatically guarantee that every segment of a community has participated equally.
A digital civic conversation could potentially reach thousands of people while still underrepresenting residents who lack reliable internet access, have language barriers or are less comfortable participating online.
Jigsaw's own guidance recognizes that recruitment and community trust are important elements of successful civic conversations. Its published framework recommends establishing the intended audience, designing outreach and recruitment strategies and considering how participants will be engaged before the AI analysis begins.
That means the AI is only one component of the broader process.
The quality of the resulting insight depends partly on who participates and what questions are asked.
This is especially important because the technology is being introduced into a politically sensitive environment without itself being designed as a political decision-maker.
Jigsaw's objective is to improve the ability of people and institutions to exchange information and understand one another. It is not an attempt to replace elected officials or establish an automated system for determining what a community should do.
The distinction could help define how AI is incorporated into civic institutions more broadly.
There are several possible roles for AI in government.
It can automate administrative work.
It can help employees search large databases.
It can translate public information.
It can analyze infrastructure and service data.
And, as Jigsaw's work demonstrates, it can potentially help governments understand large amounts of public feedback.
Each application comes with different risks and benefits.
Google has been expanding across all of these areas.
In September, Google.org announced the recipients of its $30 million AI for Government Innovation challenge, with projects ranging from emergency response coordination and government correspondence to public transportation, infrastructure management and healthcare. The recipients were selected from more than 2,600 proposals and will receive funding along with technical support from Google engineers and product specialists.
Google Public Sector has likewise described AI as an increasingly important technology for state and local government, highlighting applications involving administrative automation, data integration and service delivery.
The Jigsaw program adds another dimension to that strategy.
It focuses not simply on how governments work internally, but on how governments listen externally.
That could be one of the more consequential applications of AI because public participation is fundamentally an information problem.
Governments need to know what residents think.
Residents need meaningful ways to communicate those views.
And both sides can struggle when the volume of information becomes too large to process through traditional methods.
AI can potentially reduce that bottleneck.
The challenge will be ensuring that greater scale does not come at the expense of accuracy, privacy or human judgment.
If a city receives 10,000 comments, an AI system may make it easier to understand them.
But the resulting summary still needs to be treated as an interpretation rather than an unquestionable representation of the community.
That means human oversight remains central.
Jigsaw's model ultimately envisions AI as an intermediary that helps people understand conversations rather than as an authority that determines their outcome.
The approach is consistent with the organization's broader concept of “structural agency.” Jigsaw argues that modern society is highly connected technologically while many people can still feel disconnected from the decisions and institutions affecting their lives. Its civic AI projects are an attempt to use technology to narrow that gap.
The expansion of Sensemaking AI could therefore mark an important experiment in the next stage of government technology.
For decades, civic engagement has relied on meetings, surveys, hearings, focus groups and written comments.
AI does not eliminate those methods.
Instead, it could provide another layer capable of processing conversations at a scale that would otherwise be difficult for government staff to manage.
The results from Bowling Green and other early participants will provide an indication of how useful that approach can become as it expands.
For Google, the project also provides a practical demonstration of how its AI technology can be applied outside conventional consumer products.
Rather than asking residents to interact with an AI chatbot, the system is being positioned as infrastructure for human conversations.
The AI's job is to help organize what people say.
That could prove to be one of the more consequential directions for government AI because the technology's value would not come from replacing human participation, but from making large-scale participation easier to understand.
As the Jigsaw Partner Program expands, more cities will have the opportunity to test that model.
The experiment will ultimately raise a larger question for governments around the world: Can artificial intelligence help institutions listen to more people without reducing the complexity of what those people are saying?
Google and Jigsaw are betting that it can.
The technology now has an opportunity to prove it in communities where the most important part of the process will remain decidedly human—the people speaking, the officials listening and the decisions made after the conversation ends.