Speaking with Lorelei Kelly underscored how often debates about democratic decline focus on politics while overlooking the institutional systems that make representation possible.
Kelly situates modernization within a long constitutional tradition of structured civic voice feeding directly into lawmaking. Her work invites us to see Congress as an information system; one that has struggled to keep pace with the scale, speed, and complexity of modern governance.
What makes this conversation especially important is the role emerging technologies may play in that evolution. AI, if used innovatively in the public deliberation process, could help institutions sift through large volumes of civic input, turning participation into insight rather than noise.
If Congress is meant to metabolize the ideas, needs, and experiences of hundreds of millions of people, the tools enabling that exchange matter just as much as the politics shaping the outcomes.
In that sense, renewal is not just about modernization; it is about reimagining how citizens, communities, and institutions interact, restoring the connective tissue between communities and the institutions meant to represent them.
Interview with Lorelei Kelly, Public Good Group, to discuss newly released research: The First Amendment Promise of Deliberative Technology: Reviving Assembly and Petition to Modernize the U.S. Congress
Q: What drew you to the challenge of strengthening Congress as an institution, and how has your thinking about civic voice and democratic infrastructure evolved over time?
Lorelei: I need to go back a couple of decades. I see strong democratic infrastructure as a peace and security issue, because my life experiences are where these issues overlap.
I lived in West Berlin in 1989, researching nuclear arms control at the twilight of the Cold War, a defense posture of escalating nuclear danger between the US and the Soviet Union that put the entire world at existential risk. I became deeply involved in democratic movements across Eastern Europe—from the Soviet Union and Czechoslovakia to Hungary and East Germany—and dedicated myself to the freedoms that democratic society provides.
A decade later, I was brought into Congress by a friend who had just been elected. She asked me to work on arms control and WMD policy, and while there in the late 1990s, I picked up everything else that had expanded the security agenda: humanitarian response, pandemics, human rights, diplomacy, climate change, terrorism, financial corruption, and cyber threats.
I pivoted into democratic technology around 2012 because I realized that if we want the most powerful legislature in the world to understand complex threats and produce peaceful outcomes, we have to change the information architecture of the entire system.
Congress is the heartbeat of Article I in our Constitution for a reason; our country was founded on the concept of civic voice in democratic government. And yet Congress was still preparing for last-century threats, behaving as if waiting for Napoleon to invade Prussia.
I was working on the Hill on 9/11, and that tragedy exposed how poorly our institutions had adapted. We had moved from a world where strong states were the primary threat to one where weak states were, where the danger was no longer conventional weapons held by nation-states but the knowledge to create them, proliferating across boundaries that no longer contained it. Our leaders largely did not see it coming. The wars in Iraq and Afghanistan further illustrated our inability to handle that complexity.
Today, democratic infrastructure is more critical than ever. Congress has been hollowed out, intentionally and through decades of neglect, and it is decades behind both the private sector and the executive branch in its ability to compete on ideas or solutions. It took a pandemic to move it forward.
During COVID, Congress moved from clipboards to electronic workflows almost overnight, hearings went remote, staff became talented online moderators, and one of my favorite statistics is that diversity among witnesses increased more than 100 percent when committee hearings moved to Zoom. And yet, the support agencies of this institution only switched from the free version of Zoom last year, and it deployed enterprise Wi-Fi across district offices only in 2023.
Authentic civic voice in democratic institutions is the new frontier of representative democracy. Involving Americans in their own self-government beyond elections is essential if we want this experiment to endure.
The information system feeding Congress has largely been captured by narrow private interests, and the ecosystem of weaponized information is a massive problem. In my view, an authentic civic voice in democratic institutions is the new frontier of representative democracy. Involving Americans in their own self-government beyond elections is essential if we want this experiment to endure.
Q: Your new report argues that democratic technology is not simply a matter of modernization but of constitutional renewal. What is the core argument?
Lorelei: The central argument is that deliberative technology is not something new that needs to be invented from scratch. It is rooted in two underdeveloped rights embedded in the First Amendment: assembly and petition.
Historically, most Americans could not vote. To seek redress of grievances, they assembled and petitioned Congress in structured ways that became part of the public record. Petitioning was not symbolic; it was a workflow inside Congress. Members would accept petitions about a collective issue, introduce them, debate them, and advocate for the people who submitted them. That process drove some of the most important democratic transformations in American history. The abolition movement relied heavily on petitioning Congress. So did the movement for women's suffrage.
Congress exchanged democracy for efficiency and has never fully recovered its civic voice in its workflow.
Over time, Congress abandoned much of this civic input infrastructure. The modern committee hearing system evolved from those petitioning practices, but the connection to civic voice faded as the institution outsourced information flows into issue-specific federal agencies. In this way, Congress exchanged democracy for efficiency and has never fully recovered its civic voice in its workflow.
The report argues that if we want to modernize Congress for the twenty-first century, we need to revive those constitutional capacities using modern tools. And it is worth noting that this is not a radical or partisan argument. Our Constitution enshrined skepticism of consolidated power; federalism exists to distribute it.
The consolidation of authority into the executive branch has been building for decades and has reached a kind of escape velocity. Reviving assembly and petition as genuine governing functions is one of the most constitutionally grounded responses we have.
Q: In the report, you demonstrate that Congress's unresponsiveness to the public is less a failure of political will than a failure of institutional infrastructure. What led you to frame the problem that way?
Some document formats in Congress date to the Civil War era.
Lorelei: As someone who has worked in and around Congress for more than two decades, it hurts to see the institution attacked so frequently. The people who work there are often incredibly dedicated; many are underpaid, and they operate in a system that relies on outdated communications technology; some document formats in Congress date to the Civil War era.
The real issue is that Congress is functioning as an information intermediary between communities and national power, but it has been using tools from the last century. The United States has one of the most complex democratic systems in the world — two chambers with different cultures, rules, and operational rhythms, and unlike most institutions, no single executive leader or information authority.
There are 541 offices communicating at once, creating an overwhelming, indecipherable information environment.
There are 541 offices communicating at once, creating an overwhelming, indecipherable information environment. This has nationalized American political communication around big culture war issues and pushed local civic voices to the margins, and local is where problem-solving actually happens.
In practice, that environment has been captured by well-resourced actors who can purchase access. The 2010 Citizens United decision reinforced that dynamic by effectively declaring that money is a form of political speech, distorting the information supply chain entering policymaking. If you can buy access to lawmakers, your voice is amplified relative to everyone else.
Members are also trapped in incentives that push them toward performance rather than governing, pulled toward fundraising, media appearances, and partisan messaging because there are only so many hours in the day.
Technology could help solve what I call a Moneyball problem: helping members surface hidden ideas earlier, evaluate trade-offs, and build coalitions before policy debates become polarized. That has to happen through deliberation, and fixing the infrastructure to enable it should be an issue that candidates run on.
Q: Your report makes a distinction between civic technology and deliberative technology. What does that distinction mean in practice?
Deliberative technology focuses on helping people reason together and generate insight that institutions can actually use.
Lorelei: Civic technology typically focuses on participation, tools that allow people to contact their representatives or sign petitions. Deliberative technology focuses on helping people reason together and generate insight that institutions can actually use.
Because of Citizens United, you now need either money or people to have a voice in lawmaking, and we still organize people like it's 1980. We have to go back to basics and develop the underused capacities of our democracy. Organizing civic voice in modern ways is one of the most underused.
One example I highlight in the report is Civil Society Field Hearings, which are structured community listening processes modeled on congressional hearings themselves. This came directly out of work I did on the Hill, where I created convening mechanisms beneath the committee systems to help staff across foreign relations, science and technology, and health find each other and collaborate.
Civil Society Field Hearings apply the same principle at the community level. Communities organize them locally, document testimony in a standardized format, and produce a credible public record. This means listening, documenting, and formatting in a structured way, and making that information accessible at no cost. The goal is to create trustworthy information that Congress can metabolize.
This is built on a genuine belief in the potential of human nature: people want to solve problems and contribute to their communities. They want to be generous. Our system right now does not optimize for that generosity, and we need to build one that does.
My inspiration was the Sunflower Movement in Taiwan in 2014. Audrey Tang, one of its leaders, always says they were not only protesting but also demonstrating a new way to govern. Two global leaders on digital government are Estonia and Taiwan. Both see democracy as the social cohesion that makes a nation strong. There are also powerful innovations emerging from Kenya, Colombia, and across the Global South. We should be learning from all of them.
The concept I call institutionally adjacent is key: building prototypes close enough to the institution that it eventually sees them and says, this is helpful, this is trustworthy, this is part of my constitutional duties. That is far easier for a member to act on than a cacophony of protest. And importantly, all parties have members who genuinely want to figure out how to modernize their representation of constituents.
There is also untapped existing infrastructure. There are roughly 1,200 Federal Depository Libraries across the country, just now transitioning from physical stacks to digital. They are part of Congress– the First Branch of Government. These should also be playing a civic service role, connecting municipal data and community voice back to lawmaking. That potential is almost entirely unrealized.
Q: If someone wanted to help rebuild two-way communication between communities and Congress tomorrow, what practical steps could they take?
Lorelei: First, meet your local congressional staff and your district-level members. Build relationships. Ask them how they currently listen to constituents and what tools they need. Members and staff almost never get asked what they need; they are almost always on the receiving end of demands.
I call this a community listening assessment, and I have been refining the same assessment tool for thirteen years. Give people a practical script for that first meeting: What tools are you using? Do you know about modernization efforts happening in your own chamber? How can we start experimenting together?
Second, recognize the changing threat environment. Members of Congress are receiving roughly 300 percent more death threats than in previous decades. We have created a permissive environment for violence against elected leaders, people are not running for office, and people are leaving because of it. We have to completely rethink how civic participation happens as a result.
At the same time, the threat of synthetic information makes in-person engagement more important than ever, and we have to figure out how to scale it for House members, each of whom represents upward of 750,000 people. Senators represent entire states, which creates different needs and opportunities.
Third, create civic working groups or third-space environments, such as makerspaces or civic labs, where structured conversations about policy can take place. Makerspaces exist everywhere now, they are civically minded, and they are already building tools. I have long envisioned a series of Civil Society Field Hearings in makerspaces across the country, focused specifically on AI and our future. Those conversations, formatted correctly, could feed directly into policymaking. The infrastructure exists; we just need to use it.
Q: What role should artificial intelligence play in this ecosystem?
AI's primary role is to sort and filter the vast volume of information and civic voices so that community input becomes usable within legislative workflows.
Lorelei: AI's primary role is to sort and filter the vast volume of information and civic voices so that community input becomes usable within legislative workflows. The key is to prototype or experiment with deliberative technology yourself and in community, not to wait for institutions to lead.
Part of that means understanding data cooperatives and what a public-serving data infrastructure could actually look like. Public AI is an exciting global movement. If a data center is being built in your neighborhood, advocate for a public or municipal component.
Vermont’s legislature just introduced a bill to create cloud computing as a public utility. We need more like this. Our tax dollars are funding this infrastructure, and we should be attaching democratic demands to it. This includes Congress itself. That is how we begin to level the playing field.
The most urgent challenge right now is situational awareness, understanding what your member of Congress is actually trying to accomplish across all the roles they play and how they are connecting with constituents.
The institution is already changing its rules to incorporate AI into workflow; the Committee on House Administration is actively working on exactly this, as is the Congressional Digital Service. But the civic engagement piece is the hardest, because it is the most vulnerable to being weaponized or gamed.
That is why there is no substitute for people showing up in person and in moderated online spaces to share ideas with members in ways those members can actually use. The point is not just participation but creating a credible record. Civil Society Field Hearings and Cortico's Fora are both about generating usable data from community conversations and formalizing that into something Congress can act on. That means supporting your elected leaders in their representative roles and including them in the experimentation ahead.
Q: What would you most like to see happen next, from policymakers, civil society organizations, and technologists?
Lorelei: Read the report and explore the Civil Society Field Hearing Guide at constitutionaldefense.us; that is the most immediately actionable resource, designed to help communities begin experimenting right away.
Get to know your local representatives’ staff, learn what they know, and what they need to experiment with you. There are resources, such as the community listening interview agenda, to help.
There are many promising deliberative technology tools emerging, including Cortico's Fora, which I researched in partnership with MIT, but the ecosystem is still fragmented. One thing I would love to see is a centralized hub cataloging these tools. The Council on Technology and Social Cohesion is collecting examples of pro-social tech design.
More broadly, we need to move beyond the assumption that traditional protest and advocacy alone can shift the system. Today, almost every form of civic engagement must ultimately connect back into institutional decision-making. That is more challenging, but it is also an extraordinary opportunity given the technology and data tools at hand.
For the first time in my career, I believe we have the networks, technologies, and civic energy needed to redesign democratic infrastructure in ways that genuinely strengthen representative government.
For the first time in my career, I believe we have the networks, technologies, and the in person civic organizing energy needed to redesign democratic infrastructure in ways that genuinely strengthen representative government.
We can do this. And the place to start is at the local level.