Candidates Explore AI for Responsive Campaigning

"AI can help candidates listen more deeply to voters’ concerns, engage constituents more inclusively, and formulate policy platforms that are more responsive to our input." — the essay that originally appeared in The Guardian

Why the authors say current campaign AI is failing

The essay makes a blunt case: most AI in American campaigns so far is little more than reheated advertising. It compares present practice to "the web ads of 30 years ago, or television ads before that" — one-to-many broadcasting that inundates voters with messages rather than hearing from them. The authors argue this approach is uninspiring and avoidable: AI can magnify one-to-one connections or enable many-to-many deliberation at scale, but the dominant use in the United States has not yet embraced those possibilities.

Team Mirai’s AI interviewer: deep listening at scale

Team Mirai, a newly founded Japanese political party that calls itself a “utility party,” is offered as a practical example. The party built an AI interviewer to solicit substantive constituent input across a wide array of legislative issues and paired that capability with an AI-powered portal for exploring bills. Japanese voters have exchanged more than 300,000 messages across 16,000 AI interviews, organizers say. Team Mirai members have cited constituent AI interviews during legislative committee hearings, published syntheses of that input for voters, and amended their policy platform on the basis of what they heard — steps the essay links directly to the party’s rapid rise to win 12 seats in the Diet.

CrownShy’s Comhairle and open-source civic tools

For many-to-many deliberation, the essay highlights CrownShy, a company funded in part by the Scottish government, which is building Comhairle. That platform combines AI interviewing with synthesis of diverse viewpoints, virtual assemblies, and video testimonials to help legislatures—or campaigners—organize large-scale digital consultations. The essay emphasizes a design choice shared by Team Mirai and CrownShy: both projects are open-source and framed as tools to improve democratic processes rather than purely for partisan gain.

U.S. projects already testing deliberation tools

The essay catalogs several U.S.-grown civic technologies that have been used for public listening or policy development. Stanford-affiliated deliberation.io has been used in pilot listening sessions by the city of Washington DC. The MIT-affiliated Cortico project provides software to surface under-heard community perspectives and is organizing listening sessions at libraries across the country. The non-profit Talk to the City applies AI to analyze large datasets of stakeholder input. And the startup Remesh offers a commercial product that uses AI to generate recommendations from dialogue, tested in policy development scenarios.

What this means for candidates, technologists, and voters

• Candidates: The essay urges congressional candidates to look beyond mass messaging and to adopt responsible AI for "broad listening and deliberation" so platforms are more responsive and community-driven. It notes a political signal from Higher Ground Labs, which launched a fund this spring targeting “AI-Native Campaign Systems” and “community-Led Messaging Platforms that surface authentic, bottom-up insights from real conversations.”
• Technologists and civic innovators: The examples cited — Team Mirai, CrownShy, deliberation.io, Cortico, Talk to the City, Remesh — show multiple technical approaches (AI interviewing, synthesis, virtual assemblies) and a preference in several projects for open-source designs intended for reuse by others.
• Voters and the public: The essay acknowledges anxieties in the electorate — "voters have anxiety about AI’s impacts" — and warns that bad uses already exist, including "AI deepfakes to spread lies" and what the authors call "The White House is posting slopaganda." Those same voters may be more willing to engage, however, when they see evidence their input is being listened to and acted upon, as Team Mirai’s adoption suggests.

The authors do not argue that technology alone will fix politics. They separate tools from incentives: "We want big tech money out of politics, holding the AI companies accountable for the harm their models cause, taxing their revenues, and maybe even nationalizing them if the AI bubble bursts," the essay says, while still urging campaigns to adopt responsible AI practices. They also register a rueful institutional memory: two decades ago the civic tech surge led Congress to fund the US Digital Service — parenthetically noting, in the essay’s words, that "The Trump administration repurposed the USDS to become the US Doge Service in 2025."

Practical proof exists. Team Mirai’s 16,000 interviews and CrownShy’s Comhairle are not hypothetical designs but running experiments that have been scaled and, in Team Mirai’s case, cited in formal legislative processes. The United States already hosts several tools that perform parts of this workflow: interviewing, synthesis, assembly and recommendation. The remaining question the authors leave plainly on the table is whether American campaigns will treat AI as a louder megaphone or as a new way to listen — and whether voters will see the listening as genuine.

Read the original essay: https://www.schneier.com/blog/archives/2026/09/how-candidates-could-use-ai-for-good.html