How AI Can Deepen Democracy and Unlock Public Wisdom
In a compelling and timely essay published by the Carnegie Endowment for International Peace, author Rahmin Sarabi explores how artificial intelligence—particularly large language models (LLMs)—can play a pivotal role in reviving democratic governance through deeper public engagement and more inclusive policymaking. The article underscores how democracy in the United States is under intense strain due to polarization, distrust in institutions, and the sheer complexity of modern societal challenges. Yet, rather than replacing human judgment, AI has the potential to reinforce democratic values like participation, transparency, and collaboration, if deployed thoughtfully.
Sarabi highlights that traditional forms of civic participation, such as public forums and multiple-choice surveys, often fail to capture the nuance and breadth of community sentiment. Open-ended responses, personal reflections, and deliberative discussions provide far richer insights but have historically been underused due to the time and resources required to process them. That’s where AI comes in. With recent breakthroughs in natural language processing, LLMs from companies like OpenAI and Anthropic can now synthesize thousands of public comments, capturing the complexity of perspectives without reducing them to oversimplified talking points. These tools can unearth underlying values, highlight areas of agreement and contention, and provide decision-makers with a more human-centered understanding of their communities.
Examples from across the globe illustrate AI's growing utility in democratic engagement. In Fort Collins, Colorado, AI helped analyze over 4,000 long-form responses to a land-use issue, allowing for community-driven consensus and coordinated action. In Bowling Green, Kentucky, AI tools facilitated the inclusion of nearly 8,000 voices in long-term community planning. At the national level, France employed AI during its Citizens’ Convention on the End of Life, enabling citizens to query the recommendations in plain language and making complex deliberative outcomes more accessible. Back in California, the Engaged California initiative used AI to analyze over 1,000 responses following devastating wildfires, yielding actionable insights while preserving the authenticity of resident voices.
Despite these successes, Sarabi wisely warns against uncritical optimism. The essay underscores the real and pressing risks that AI poses to democratic integrity. Data privacy and ownership, particularly when civic engagement relies on tools developed by private companies, must be taken seriously. California’s CalCompute initiative and New York’s Empire AI consortium are positive examples of public investment in AI infrastructure designed to serve democratic, not corporate, ends. The author also cautions against the dangers of algorithmic bias, hallucination, and so-called “participation-washing,” where AI-enhanced public engagement is used as a facade for genuine inclusion without influencing actual decisions.
Most crucially, Sarabi affirms that human expertise must remain at the heart of AI-assisted democratic processes. Skilled facilitators, with their ability to read group dynamics, apply cultural sensitivity, and adapt in real time, cannot be replaced by even the most advanced AI systems. The goal is not a democracy run by machines but one enriched by them—where AI empowers human decision-makers and citizens alike to collaborate more effectively.
Looking ahead, the essay outlines a vision for 2030 in which AI-enhanced civic engagement becomes as commonplace as online voter registration—trusted, inclusive, and impactful. Achieving this vision will require deliberate investment in pilot programs, public infrastructure, and rigorous evaluation standards to ensure that AI genuinely strengthens democratic outcomes. The window to act is limited, and the stakes are high. Whether AI becomes a tool for deepening democracy or exacerbating its challenges will depend on the choices we make now.
Disclaimer: This blog post is a summary of the article “How AI Can Unlock Public Wisdom and Revitalize Democratic Governance” by Rahmin Sarabi, published by the Carnegie Endowment for International Peace on July 22, 2025. The views expressed herein are those of the original author and do not necessarily reflect the views of Carnegie or its affiliates. This summary is provided for informational purposes only and should not be construed as legal, policy, or investment advice.