Abstract

Excerpted From: Spencer Overton, Analyzing the Benefits of Artificial Intelligence to Racially Inclusive Democracy, 2026 Utah L. Rev. 541 (2026) (Full Document) (Footnotes omitted.)

Although demographic trends suggest the United States will have no majority ethnic group by the year 2050, election law is evolving in ways that obstruct truly representative democracy. State legislatures have enacted a record number of voting restrictions, often disproportionately affecting voters of color. An increasing number of U.S. Supreme Court justices have sought to scale back voting rights protections, enabling entrenched politicians to gerrymander districts and implement election rules that constrain the political influence of communities of color.

Recently, some artificial intelligence (“AI”) optimists have asserted that advancements in AI could “enhance our democracy in a way that strengthens our collective ability to govern.” They argue that AI could facilitate citizen engagement with elected officials, allow for “a more inclusive, participatory, and deliberative form of democracy,” improve collective decision-making, and perhaps even replace legislators to better represent constituents. Human beings in government repeatedly make decisions that are biased, and AI can lead to fairer enforcement and more efficient administration of government benefits.

Unfortunately, these AI optimists largely overlook one of the most pressing challenges to democracy in our increasingly diverse nation: the legal gaps created by the Court’s rollback of voting rights protections, which have made it easier for politicians to manipulate voting rules. Like the technologies that AI optimists champion, mainstream AI-optimist literature often lacks cultural nuance and fails to grapple with race and democracy, and this Article aims to provide that missing lens. Despite race being the most significant demographic factor that shapes voting patterns in the United States--and despite widening racial disparities in voter turnout--the work of AI optimists remains largely oblivious to race. Communities of color face longer lines at the polls than whites and are underrepresented in elected office, yet AI optimists rarely acknowledge these issues. They also fail to critically examine how their proposed AI applications and legal proposals might exacerbate racial disparities in political power.

This Article addresses the gaps in both legal protections and the AI democracy literature, offering the first comprehensive analysis of the potential benefits of AI and AI law in mitigating democracy’s racial challenges. Most literature on the democracy-enhancing uses of AI falls into an effective-altruism trap that has huge blind spots with regard to economic, political, and racial incentives. In contrast, this Article offers a clear-eyed assessment of real-world impediments to deploying AI to make democracy more racially inclusive. It then evaluates the effectiveness of various AI applications in solving democracy’s racial challenges. In exploring AI’s potential benefits, this Article complements my earlier work on AI’s potential harms to racially inclusive democracy.

AI is not a magic wand that will erase democracy’s most entrenched racial and political problems, but it also need not reinforce the racially homogenous system of governance envisioned by many of the framers of the U.S. Constitution. The assumption that AI will inevitably lead to either a democratic utopia or an authoritarian dystopia cedes too much power to machines and absolves humans of agency and accountability. Even accepting that AI tends to exacerbate “existing power imbalances” among humans, one need not embrace techno-utopianism to accept that some AI applications could benefit racially inclusive democracy.

For example, community organizations could leverage AI to identify and engage low-turnout voters, reducing racial disparities in electoral participation. Community organizations could increase government responsiveness by using AI to understand their constituents’ priorities, track legislative activity, analyze policies, and assist constituents in submitting public comments on key issues. Election administrators could deploy AI to allocate polling equipment and workers more equitably, combat disinformation targeting communities of color, and detect voter suppression efforts. Civil rights organizations could use AI to more effectively detect and challenge discriminatory voting restrictions, electoral districts that dilute minority voting strength, and other civil rights violations.

This Article asserts that a central goal of AI and the law that regulates it should be to facilitate a well-functioning, racially inclusive, pluralistic democracy--one that respects both identity and individual autonomy and enables cross-group engagement and coalition building. AI tools and laws should advance principles like widespread participation, overcoming racial polarization, and government responsiveness.

AI law is no substitute for an updated Voting Rights Act and a Supreme Court that respects the 15th Amendment. AI law will play a critical role, however, in determining whether AI is deployed in ways that advance racially inclusive democracy. If left unchecked, profit-driven AI development will likely prioritize those who have substantial financial resources, further exacerbating existing racial disparities in democracy. The most pressing threat is not a hypothetical, all-powerful AI ruling over humanity, but rather elite humans leveraging their control over technology to subjugate those who lack such power.

To counter this, lawmakers must act now to ensure AI is developed and deployed in ways that promote racially inclusive democracy. This includes expanding access to trustworthy AI--particularly in regions that have been economically and politically marginalized, such as Tribal Lands, the Colonias of the Southwest, the Black Rural South, the U.S. Territories, Central Appalachia, and high-poverty, low-voter-turnout census tracts. Federal, state, and local laws should prioritize widespread access to broadband, digital-skills training, computers, and public AI to empower a racially diverse citizenry. Because many will rely on private AI, robust legal safeguards are necessary that mitigate bias, prohibit manipulation, and require that companies only collect and retain the personal data strictly needed to achieve a specific purpose (“data minimization”).

This Article proceeds in four Parts. Part I defines AI in the context of democracy and details the shortcomings of AI optimists’ arguments. Part II contends that human values and decision-making--rather than AI--will ultimately shape the future of American democracy. It introduces three core principles for a racially inclusive, pluralistic democracy: (1) widespread participation; (2) social connectedness to manage the harms of racial polarization; and (3) government responsiveness to diverse constituencies. Part III assesses the likely effectiveness of various AI applications in advancing these three principles and mitigating racial challenges to them. Finally, Part IV discusses legal solutions, including AI accountability laws and widespread access to public AI infrastructure, so that America’s increasingly diverse population can harness beneficial applications of AI and more fully engage in democracy.

[ . . . ]

AI optimists envision a future in which technology enhances democracy and reduces societal inequality, but the benefits of AI are not inevitable. Supporters of racially inclusive democracy need to push for AI law and discrete technologies that are carefully tailored to advance three key democratic principles: (1) widespread participation; (2) connectedness; and (3) government responsiveness. AI should serve as a supplement to, not a substitute for, human judgment and collective deliberation--enhancing the capacity of people, groups, and institutions to act inclusively and effectively rather than completely automating decisions and activities that demand human accountability and moral reasoning. By advancing these principles, AI law has the potential to shape the trajectory of our democracy by mitigating challenges like racial disparities in voter turnout, racially polarized voting, and inequitable governance.

Patricia Roberts Harris Research Professor and Founder and Faculty Director of the GW Multiracial Democracy Project, The George Washington University Law School.