Artificial Intelligence and the Automated Ballot: Navigating Assistance, Autonomy, and the Future of Democratic Choice

1. Introduction: The Voter’s Dilemma and the Allure of AI Assistance

The modern election ballot often presents citizens with a formidable array of choices. From candidates for numerous offices to complex referenda and judicial retention questions, the sheer volume of decisions can lead to voter fatigue or, worse, uninformed selections. This “voter information deficit” is a recognized challenge in contemporary democracies, where citizens may struggle to access or process the information needed to make choices aligned with their interests and values.¹ In an era characterized by rapid advancements in artificial intelligence (AI), it is a natural inclination to look towards technological solutions to simplify such complex tasks. The prospect of AI offering a streamlined path through the intricacies of the ballot may seem appealing to voters grappling with information overload.

This report investigates the current landscape of AI-based tools in relation to election processes, with a specific focus on the query of whether tools exist that can automatically fill out election ballots for citizens. It will analyze the technical feasibility of such systems, explore the existing legal and regulatory frameworks pertinent to voter assistance and AI, and delve into the profound ethical implications and democratic risks associated with the concept of AI-driven ballot completion. Ultimately, the report aims to provide clarity on the current state of AI in elections and to discuss the principles that should guide the responsible use of AI in the broader civic sphere.

While AI holds considerable promise for enhancing voter information and engagement, the evidence indicates that tools designed to automatically fill out official election ballots for citizens do not currently exist within legitimate democratic processes. Furthermore, the development and deployment of such technologies would present formidable and potentially insurmountable legal, ethical, and democratic challenges, fundamentally altering the nature of civic participation and the integrity of the electoral process. The user’s query itself, while seemingly focused on a technological solution, touches upon a deeper societal inclination towards efficiency and a potential techno-solutionist approach to complex civic duties. The desire to delegate a cognitive and civic task like voting to an AI reflects a broader trend where technology is increasingly seen as a means to optimize or offload responsibilities. This raises questions about whether the act of personal deliberation and choice in voting might be undervalued in pursuit of convenience, potentially reshaping our understanding of civic responsibility. The very idea of AI providing “help to decide” sits precariously between beneficial assistance and undue influence, a boundary that AI could easily blur if not meticulously designed and rigorously regulated.

2. AI in the Election Sphere: Current Applications and the Quest for Voter Support

The integration of artificial intelligence into the election sphere is multifaceted, with current applications primarily falling into categories that support election administration, voter information dissemination, and, more controversially, political campaigning. However, these applications are distinct from the concept of an AI tool that would autonomously complete a voter’s official ballot.

AI in Election Administration:

Several technological tools, some incorporating AI principles, are employed by election officials or organizations to manage the mechanics of elections. For instance, platforms like Election Runner 2 and ElectionBuddy 3 offer services for creating and managing secure online elections, primarily for schools, unions, and other private organizations. These systems focus on aspects like voter ID and key management, mobile accessibility, ballot creation interfaces for administrators, and secure vote tallying.2 Similarly, solutions such as BlueCrest’s Relia-Vote Ballot Manager are designed to streamline the vote-by-mail and absentee ballot process for election officials, covering production, distribution, return, and validation, with an emphasis on mailpiece tracking and integrity.4 The AES AutoVote system functions as an electronic poll book and on-demand ballot printing system, primarily used to manage voter check-in and ballot issuance in specific jurisdictions.5 These administrative tools enhance the efficiency and security of the election process but do not engage in making substantive choices on behalf of voters in public elections.

AI for Voter Information and Engagement:

A more direct line of AI application relevant to voter assistance involves tools designed to inform and engage citizens. Research initiatives, such as a project at the University of Virginia, aim to leverage AI to make primary-source material like proposed bills and legislative histories “much more digestible and relevant to individual voters”.1 The goal is to address the “voter information deficit” by enhancing understanding, not by automating choices. Notably, early assessments within this project found that general AI models like ChatGPT provide generic information and are “unwilling to provide an answer” when directly asked, “Which candidate best represents me?”.1

A practical example is GoVoteBot, an AI-powered chatbot developed to help voters navigate procedural complexities such as voter registration, locating polling places, and understanding election deadlines.⁶ This tool is explicitly non-partisan and “has no opinion on whom people vote for,” focusing solely on facilitating the act of voting.⁶ Established non-partisan organizations like Ballotpedia ⁷, Vote Smart ⁹, and the League of Women Voters (through its VOTE411 platform ¹¹) provide extensive factual data, candidate stances, and explanations of ballot measures. While they utilize technology to disseminate this information, their core mission is to empower voters to make their own informed decisions, not to employ AI to recommend choices or complete ballots.

AI in Political Campaigning and Persuasion:

A distinct and often concerning application of AI in the electoral domain is its use in political campaigning. AI is employed for micro-targeting voters with tailored messages, generating campaign materials, and, in some instances, creating synthetic media, or “deepfakes”.13 Research from institutions like the University of Chicago and Stanford University highlights concerns about “AI-powered micro-targeting or emotionally manipulative chatbots” and the potential for AI to be used to create and disseminate “false or misleading information”.13 The National Conference of State Legislatures (NCSL) reports that, as of May 2025, twenty-four states have enacted laws to regulate political deepfakes, underscoring the perceived threat of AI-generated deceptive content in campaigns.14 These uses of AI are designed to influence voters, rather than to assist them in an objective, voter-initiated manner.

The current landscape of AI in elections thus reveals a clear divergence. Tools for election mechanics and for voter information dissemination are evolving, but a significant gap exists before reaching AI for automated personal ballot completion in public elections. This gap is not merely a technological hurdle; it is deeply rooted in normative and legal principles. Automating a ballot choice is fundamentally different from managing election logistics or providing neutral information; it involves the delegation of personal agency, intent, and legal standing in the sacrosanct act of voting. The substantial legal and ethical considerations, discussed later in this report, explain why this chasm persists despite AI’s advancing capabilities in other decision-support domains.

Furthermore, the proliferation of AI in campaign persuasion, particularly the rise of deepfakes and sophisticated micro-targeting, creates a reactive pressure for enhanced AI literacy and critical assessment tools for voters. The threat of manipulative AI paradoxically underscores the need for strengthened human agency and critical judgment, rather than its delegation to another AI system. If voters are already contending with AI-driven attempts to sway their opinions, introducing an AI to make choices for them could render them more vulnerable, especially if the decision-making AI itself is flawed, biased, or compromised. Thus, the negative uses of AI in campaigning serve as a compelling argument against AI ballot-filling by highlighting the paramount importance of preserving and fortifying the voter’s own capacity for critical decision-making.

To clarify these distinctions, the following table outlines the spectrum of AI involvement in elections:

Table 1: Spectrum of AI Involvement in Elections

3. Automated Ballot Completion by AI: Reality, Feasibility, and Inherent Limitations

Based on extensive review of available research and current practices, no publicly acknowledged or legally sanctioned AI tools exist that automatically fill out official election ballots for citizens in democratic elections. While platforms like Election Runner ² and ElectionBuddy ³ facilitate the creation and administration of online elections for private organizations, and systems like BlueCrest’s Relia-Vote Ballot Manager ⁴ and AES AutoVote ⁵ assist election officials with ballot processing and poll book management, these do not involve AI making substantive choices on behalf of individual voters in public electoral contests.

Technical Feasibility (Hypothetical):

Theoretically, one could envision an AI tool designed to assist with ballot completion. Such a system might function by having a user input their policy preferences, ethical values, demographic information, and perhaps even their sentiments on various issues. The AI would then attempt to match this profile against available data on candidates’ platforms, past voting records, public statements, and the specifics of ballot measures. AI’s known capabilities in pattern matching and synthesizing large volumes of information could, in principle, be applied to such a task.1

Inherent Limitations and Challenges:

Despite the theoretical possibility, the practical realization of a reliable and ethical AI ballot-filler faces profound limitations:

• Nuance of Human Preferences: Political views are rarely simple or unidimensional. They are often deeply nuanced, can hold apparent contradictions, are highly context-dependent, and are frequently intertwined with emotions and personal experiences. Current AI, including sophisticated large language models, struggles to capture and adequately process this level of human complexity. As noted by researchers at the University of Virginia, when students asked ChatGPT, “Which candidate best represents me?”, the program was “unwilling to provide an answer,” and the information it did provide was described as “generic” and reflective of “candidate’s talking points”.¹ This suggests a significant gap between current AI capabilities and the ability to make personalized political recommendations suitable for direct ballot completion. The challenge here is not merely about processing power or data access; it is a fundamental problem of translating deeply human, often non-quantifiable, and evolving political identities and preferences into algorithmic logic. Voting decisions are informed by an array of factors including values, emotions, trust, identity, and reactions to dynamic current events, not just static policy preferences. The attempt to codify the essence of individual political judgment into machine-readable format represents a qualitative, not just quantitative, hurdle.
• Data Availability and Bias: The recommendations generated by such an AI would be entirely dependent on the data it is trained on and has access to—candidate statements, media coverage, voting records, policy analyses, etc. This data can be incomplete, strategically framed by political actors, or inherently biased. If the underlying data reflects societal biases (e.g., in media representation of candidates or issues), the AI will inevitably perpetuate and potentially amplify these biases in its suggestions.¹⁵ The U.S. Election Assistance Commission (EAC) warns that AI can “accelerate false or biased information” ¹⁶, a risk that becomes acute if biased AI is directly influencing votes.
• Dynamic Nature of Politics: The political landscape is constantly shifting. Candidates may alter their positions, new issues can emerge rapidly, and unforeseen events can reshape voter priorities. Maintaining an AI’s knowledge base to be consistently current, comprehensive, and unbiased across all relevant electoral contests would be a monumental and continuous undertaking.
• Explainability and Trust: A critical component of any AI system, especially one involved in sensitive decisions, is transparency and explainability. How would an AI ballot-filler articulate the rationale behind its choices in a way that is understandable and verifiable by the voter? A lack of transparency can quickly erode trust.¹ If the AI operates as a “black box,” voters may be unable to ascertain whether its recommendations genuinely align with their nuanced intentions or if they are the product of flawed logic or hidden biases. An AI ballot-filling tool, even if technically refined, could thus create a “black box” voting experience for many users. This would undermine the educational and civic engagement aspects inherent in the voting process, potentially leading to a less informed and more passive citizenry. If voters merely accept AI-generated choices without engaging with the underlying issues or candidate platforms, their ability to understand political discourse and hold elected officials accountable could diminish.
• Defining the “Best Choice”: Politics often involves trade-offs between competing values or desirable outcomes. Whose definition of the “best choice” would the AI employ? How would it weigh conflicting priorities for a single voter—for example, a desire for lower taxes versus stronger environmental protections, if no single candidate perfectly embodies both? The attempt to optimize a political choice algorithmically is fraught with subjective judgments that must be programmed into the system, often opaquely.

AI systems, particularly those designed to support decision-making by mitigating uncertainty, carry inherent risks.¹⁵ Concerns about “algorithmic biases, lack in training data, and flawed AI models,” as well as the potential for AI to create “false illusions of (absolute) certainty,” are highly relevant to the concept of an AI ballot-filler.¹⁵

4. Legal and Regulatory Landscape: Voter Assistance, AI, and Election Integrity

The legal framework governing elections in the United States, particularly concerning voter assistance, is built upon principles of ensuring access, preventing undue influence, and maintaining the integrity of the individual’s vote. The introduction of an AI tool to automatically fill out ballots would intersect with this framework in complex and largely unaddressed ways.

Current Voter Assistance Laws (U.S. Focus):

Federal laws such as the Voting Rights Act of 1965 (VRA) and the Help America Vote Act of 2002 (HAVA) establish provisions for voter assistance. Section 208 of the VRA, for example, stipulates that any voter who requires assistance to vote by reason of blindness, disability, or inability to read or write may be given assistance by a person of the voter’s choice, other than the voter’s employer or agent of that employer or officer or agent of the voter’s union.18 HAVA further mandates that polling places for federal elections have at least one voting system accessible to individuals with disabilities, allowing them to vote privately and independently.19

State laws often mirror these federal provisions. For instance, Ohio law permits a voter who cannot mark their ballot or needs assistance due to a disability to bring someone to help or to request assistance from two poll workers of different major political parties. Similar to federal law, there are restrictions on who can provide this assistance (e.g., not a candidate on the ballot, employer, or union agent).²⁰ These laws clearly envision human assistance, aimed at enabling the voter to cast their own vote according to their own expressed intentions, rather than delegating the decision-making process itself. The existing legal framework for voter assistance is predicated on human interaction and the explicit, contemporaneous expression of the voter’s will to a human assistant. An AI autonomously filling a ballot based on a pre-compiled profile or algorithmically derived preferences fundamentally disrupts this model. It introduces a non-human agent and choices that may have been determined by an algorithm well before the moment of voting, based on data and logic the voter may not fully comprehend. This creates a legal vacuum at best, and at worst, a direct conflict with the intent of voter assistance laws, which are designed to facilitate the voter’s own choice, not to substitute it with an external one.

Absence of Laws for AI Ballot Completion:

Crucially, no current federal or state election laws explicitly authorize, regulate, or even contemplate an AI tool autonomously marking a voter’s official ballot based on preference inputs. The legal architecture is designed for human voters and, where necessary, human assistants.

Deepfake and AI Content Regulation:

The emerging legislative activity concerning AI in elections is primarily focused on mitigating potential harms, particularly the spread of deceptive AI-generated campaign content. As of May 2025, twenty-four states have enacted laws regulating “political deepfakes”.14 These laws generally require disclosures on media indicating AI manipulation or, in a few states like Minnesota and Texas, prohibit the publication of such deepfakes close to an election.14 This legislative focus indicates that lawmakers’ current concern is centered on AI’s potential for misinformation and deception in the campaign environment, rather than its role as a direct tool for ballot completion. The fact that state-level efforts to regulate AI in elections are currently focused on these negative externalities suggests that proactive legislation for hypothetical positive uses, such as AI ballot assistance, is significantly lagging. This implies either a lack of perceived immediate need for such tools or, more likely, considerable apprehension from lawmakers about their implications for democratic processes.

Implications for Ballot Secrecy and Undue Influence:

The principle of a secret ballot is a cornerstone of democratic elections, protecting voters from coercion and intimidation. It is unclear how an AI tool, which would necessarily process a voter’s detailed preferences and then “mark” a ballot (even a digital one), could guarantee this secrecy, especially concerning the data shared with the AI and its potential storage or transmission. Furthermore, the potential for the AI itself—or more accurately, its developers, controllers, or those who influence its training data—to exert undue influence over the voter’s choices is a paramount concern. This would contravene the spirit and letter of election laws designed to ensure that the vote cast is a free and genuine expression of the voter’s will.

The following table contrasts current legal provisions for voter assistance with the implications of an AI ballot-filling tool:

Table 2: Voter Assistance: Current Legal Provisions vs. AI Intervention

5. Ethical Minefields: The Risks of AI-Driven Ballot Choices to Democratic Principles

The prospect of AI tools automatically completing election ballots, while potentially appealing from a convenience standpoint, navigates a minefield of ethical concerns that strike at the heart of democratic principles. These risks extend beyond mere technical flaws to encompass fundamental questions of autonomy, fairness, manipulation, and the very integrity of the electoral process.

• Erosion of Voter Autonomy and Agency: The act of voting is a deeply personal expression of civic duty, individual preference, and political agency. Delegating the final decision-making process to an AI, however sophisticated or well-intentioned, inherently diminishes the voter’s direct involvement and conscious endorsement of the choices made in their name. This offloading of responsibility could transform a vital act of democratic participation into a passive exercise.
• Bias and Discrimination: AI systems are not inherently objective; they learn from the data they are trained on and reflect the biases embedded within that data or their underlying algorithms.¹⁵ If an AI ballot-filler is trained on skewed datasets (e.g., media coverage that disproportionately favors certain viewpoints, or historical data reflecting discriminatory patterns), it could systematically generate recommendations that disadvantage specific candidates, parties, or demographic groups. The U.S. Election Assistance Commission (EAC) has explicitly warned that AI can “accelerate false or biased information”.¹⁶ When such biases translate directly into cast votes, the discriminatory impact is immediate and profound. Research from the MIT Civic Data Design Lab also highlights concerns about “algorithmic bias” and the risk that “biases in large language models (LLMs) could reinforce dominant cultural narratives”.²¹
• Manipulation and Undue Influence: The potential for manipulation, both intentional and unintentional, is a significant ethical hazard. Malicious actors could seek to influence an AI ballot-filler’s recommendations by “poisoning” its training data or exploiting algorithmic vulnerabilities. Even without direct malice, the persuasive capabilities of AI are a concern. Studies have pointed to the risk of “AI-powered micro-targeting or emotionally manipulative chatbots” persuading voters to act contrary to their genuine interests or polarizing the electorate.¹³ An AI tool that directly completes a ballot could become the ultimate instrument of such manipulation, subtly guiding choices based on programmed objectives unknown to the voter.
• Lack of Transparency and Accountability (The “Black Box” Problem): If an AI system recommends or, more critically, makes ballot choices, determining accountability for flawed, biased, or harmful outcomes becomes exceedingly difficult. Is the AI itself responsible? Its developers? The entity that supplied its training data? Or the voter who chose to rely on it? This “double delegation problem,” where political actors (or in this case, voters) defer decision-making to algorithms, blurs lines of responsibility.¹⁵ “Political machines,” or AI systems in governance, often operate with limited public scrutiny, and the call for transparency and explainability is paramount.¹⁷ Without clear insight into how an AI arrives at its decisions, voters cannot truly trust or verify its outputs. The introduction of AI ballot-fillers could paradoxically increase the cognitive load for conscientious voters. Instead of simply evaluating candidates and issues, they would now bear the additional burden of vetting the AI tool itself—scrutinizing its data sources, its potential biases, its security protocols, and the transparency of its algorithms. This “meta-cognitive load” could be as, or even more, complex than navigating the ballot directly, thereby rendering the promise of simplification illusory for those who take their civic duty seriously.
• Security Risks: An AI system with the capability to influence or directly complete ballots would represent an extraordinarily high-value target for cyberattacks. A compromised system could lead to large-scale, potentially undetectable vote manipulation, thereby catastrophically undermining election integrity. The EAC has noted that AI tools “may allow existing threats to scale more quickly and effectively”.¹⁶
• Impact on an Informed Citizenry: If voters increasingly rely on AI to make their choices, their personal engagement with and understanding of political issues, candidate platforms, and the implications of ballot measures may decline. This could weaken the foundation of an informed electorate, which is essential for a healthy democracy, and relates to the earlier point about the “black box” voting experience potentially fostering a more passive citizenry.
• Fairness and Equity of Access: The development and deployment of AI ballot-filling tools would raise serious questions about equitable access. Would such tools be equally available and effective for all citizens, regardless of their technological literacy, socioeconomic status, or access to digital devices? There is a risk that such technologies could exacerbate existing digital divides and create new forms of inequality in political participation.²¹

Widespread adoption of AI ballot-fillers could precipitate a fundamental shift in the locus of political power. Instead of residing with publicly accountable candidates, political parties, and the voters themselves, power could migrate towards largely unaccountable AI developers, the entities controlling the AI’s training data, and the designers of its algorithms.¹³ This represents a profound challenge to established democratic governance structures, potentially leading to a form of “algocratic governance” where pivotal political decisions are effectively shaped by opaque algorithmic processes, thereby undermining the core principles of representative democracy.

The following table summarizes the key risks associated with AI in ballot decision-making:

Table 3: Analysis of Risks Associated with AI in Ballot Decision-Making

6. The Path Forward: Responsible AI for Empowered and Informed Voters

The complexity of modern ballots is a genuine concern for many voters. However, the solution does not lie in AI systems that automatically complete ballots, thereby usurping the voter’s fundamental role. Instead, the path forward involves harnessing AI responsibly to empower and inform voters, ensuring that human agency and democratic principles remain paramount.

Focus on AI for Voter Empowerment, Not Ballot Completion:

The true potential of AI in the civic sphere lies in its ability to enhance voter understanding and access to information. AI can be a powerful tool for:

• Creating Better Informational Resources: AI can assist in summarizing complex ballot measures into clear, understandable language, providing neutral and comprehensive information about candidates and their platforms, and making voluminous legislative records or policy documents more accessible to the average citizen.¹ The University of Virginia project, for instance, explicitly aims to use AI to gather primary-source material and make it “much more digestible and relevant to individual voters”.¹
• Facilitating Procedural Navigation: AI-powered chatbots, like GoVoteBot, can effectively guide voters through the often-confusing mechanics of voting, such as registration processes, finding polling locations, and understanding deadlines for absentee ballots.⁶

Principles for Responsible AI in Civic Technology:

The development and deployment of any AI tool intended for voter assistance must be guided by stringent ethical principles. Drawing from research and expert recommendations, these should include:

• Human-in-the-Loop and People-Powered AI: Human oversight, judgment, and ultimate control are non-negotiable. As emphasized by researchers at MIT, a “people-powered Gen AI” approach requires humans to validate outputs, mitigate errors, contextualize results, and build trust.²¹ This ensures that AI augments, rather than replaces, human decision-making.
• Transparency and Explainability: Voters using AI-assisted informational tools have a right to understand how the AI processes information and arrives at the summaries or guidance it provides.¹ This includes clarity about data sources and the methodologies used.
• Bias Detection and Mitigation: Continuous and rigorous efforts are needed to identify, assess, and mitigate biases within AI algorithms and the datasets they are trained on.¹⁵ This is crucial to prevent the perpetuation of unfair or discriminatory information.
• Data Privacy and Security: Protecting the sensitive data of users who interact with civic AI tools is paramount. Robust security measures and clear privacy policies are essential.
• Non-Partisanship and Objectivity: Any AI tool designed to provide voter information must be demonstrably non-partisan and objective in its presentation of facts and analysis, akin to the pledge made by organizations like Vote Smart.⁹
• Accessibility and Digital Equity: AI-driven voter resources should be designed to be accessible and useful to all voters, taking into account varying levels of digital literacy and ensuring they do not exacerbate existing digital divides.²¹

The development of ethical AI for voter information can serve as a crucial positive counterbalance to the alarming rise of manipulative AI in political campaigns. By equipping voters with tools that enhance their understanding and critical thinking, society can foster a more informed and resilient electorate, better prepared to identify and resist disinformation, rather than a more passive one susceptible to algorithmic persuasion.

Role of Civic Education and Media Literacy:

Beyond technological solutions, strengthening civic education and media literacy is vital. Voters need the skills to critically evaluate all sources of information, including AI-generated content. Election bodies like the EAC play a role by providing resources to help election officials direct voters towards official and verified sources of election information.16

Alternative Voter Aids:

Voters seeking assistance with complex ballots already have access to a range of reliable, non-AI-driven resources:

• Official Voter Guides: Many states provide official voter information pamphlets, often produced by the Secretary of State’s office, which explain ballot measures and list candidates.
• Non-Partisan Organizations: Groups like the League of Women Voters (with its VOTE411.org platform ¹¹), Ballotpedia ⁷, and Vote Smart ⁹ offer extensive, unbiased information on candidates, their stances, voting records, and ballot initiatives.
• Sample Ballots: Reviewing a sample ballot before an election allows voters to familiarize themselves with the choices and conduct research at their own pace.

Long-term Civic Design:

Finally, consideration should be given to the design of ballots and election materials themselves. Efforts to use clearer language, better formatting, and more intuitive layouts could help reduce voter confusion and diminish the perceived need for external automated interventions.

A focus on “people-powered Gen AI” ²¹ and the imperative for “meaningful human control” ¹⁵ within the civic sphere signals a broader societal negotiation about the appropriate roles for artificial intelligence in democratic processes. This conversation extends beyond just elections. The way society approaches the integration of AI in voting—a domain with exceptionally high stakes for determining leadership and policy direction—could establish important precedents for the deployment of AI in other areas of governance. If AI were permitted to assume core democratic functions like ballot completion without robust safeguards and the preservation of human agency, it might normalize the delegation of other critical civic and governmental responsibilities to opaque algorithmic systems. Conversely, by establishing strong ethical frameworks and prioritizing human control for AI in elections, society can develop a model for responsible AI adoption across the public sector, ensuring that technology serves and enhances, rather than subverts, fundamental democratic principles.

7. Conclusion: Balancing Technological Innovation with the Sanctity of the Vote

The inquiry into whether AI-based tools can automatically fill out election ballots for citizens reveals a critical tension between the allure of technological convenience and the foundational principles of democratic participation. The analysis presented in this report leads to a clear conclusion: no such tools currently exist in legitimate democratic systems, and their potential development and deployment are fraught with profound legal, ethical, and democratic perils.

The desire to simplify the often-daunting task of navigating a complex ballot is understandable. However, the act of voting is far more than a mere selection process; it is a fundamental right, a civic responsibility, and a personal expression of an individual’s judgment and values. To delegate this act to an AI, however sophisticated, would risk eroding voter autonomy, introducing insidious biases, opening new avenues for manipulation, obscuring accountability, and ultimately diminishing the informed engagement of the citizenry. The existing legal frameworks for voter assistance are designed to empower individuals to cast their own votes, not to have those choices made by a non-human agent.

While AI-driven ballot completion is a problematic and ill-advised proposition, this does not negate the potential for AI to make positive contributions to the electoral landscape. When developed and deployed responsibly, with robust ethical safeguards and a commitment to human agency, AI can be a powerful ally in voter education and engagement. Tools that make complex information more accessible, help voters understand policy implications, or streamline the procedural aspects of voting can indeed empower citizens and enrich the democratic process.

The challenge, therefore, is not to reject technological innovation outright, but to guide it with wisdom and a steadfast commitment to democratic values. The sanctity of each individual’s vote—cast with conscious thought and free from undue influence—is paramount. As society continues to grapple with the transformative power of artificial intelligence, it is crucial to ensure that technology serves to enhance, rather than diminish, the meaning and integrity of this most fundamental act of democratic citizenship. The user’s query, while focused on a specific technological solution, ultimately prompts a vital re-examination of what voting signifies in a democracy and the non-negotiable elements of that civic act, such as individual deliberation, responsibility, and agency. The “problem” of ballot complexity, rather than being a flaw to be automated away, may also be viewed as a feature that, when appropriately supported by informational tools, encourages deeper civic engagement.

The conversation surrounding AI in elections serves as a microcosm of the broader societal challenge: integrating powerful new technologies with established human values and institutions. The choices made regarding AI in this uniquely sensitive domain will inevitably reflect and shape wider attitudes towards automation, individual autonomy, and the future of human decision-making in an increasingly technologically mediated world.

Works cited

1. Addressing the Voter Information Deficit With AI — Karsh Institute of …, accessed June 8, 2025, https://karshinstitute.virginia.edu/news/addressing-voter-information-deficit-ai
2. Election Runner: Build a Secure Online Election for Free, accessed June 8, 2025, https://electionrunner.com/
3. Voting Online, Simplified with ElectionBuddy – ElectionBuddy, accessed June 8, 2025, https://electionbuddy.com/
4. Ballot Manager Vote-By-Mail Software for Relia-Vote System – BlueCrest, accessed June 8, 2025, https://www.bluecrestinc.com/products/software/ballot-manager/
5. AES AutoVote – Verified Voting, accessed June 8, 2025, https://verifiedvoting.org/election-system/aes-autovote/
6. U.S. Citizens: Start with a Bot and End with a Ballot | U.S. Vote Foundation, accessed June 8, 2025, https://www.usvotefoundation.org/GoVoteBot
7. Official voter guides to 2024 statewide ballot measures – Ballotpedia, accessed June 8, 2025, https://ballotpedia.org/Official_voter_guides_to_2024_statewide_ballot_measures
8. Ballotpedia.org, accessed June 8, 2025, https://ballotpedia.org/Main_Page
9. About Vote Smart, accessed June 8, 2025, https://www.votesmart.org/about
10. Vote Smart, accessed June 8, 2025, https://www.votesmart.org/
11. Elections – League of Women Voters, accessed June 8, 2025, https://www.lwv.org/elections
12. Voters Guide – The League of Women Voters of Texas, accessed June 8, 2025, https://lwvtx.clubexpress.com/voters-guide
13. Preparing for Generative AI in the 2024 Election: Recommendations and Best Practices Based on Academic Research, accessed June 8, 2025, https://harris.uchicago.edu/files/ai_and_elections_best_practices_no_embargo.pdf
14. Artificial Intelligence (AI) in Elections and Campaigns, accessed June 8, 2025, https://www.ncsl.org/elections-and-campaigns/artificial-intelligence-ai-in-elections-and-campaigns
15. Opportunities and challenges of AI-systems in political … – Frontiers, accessed June 8, 2025, https://www.frontiersin.org/journals/political-science/articles/10.3389/fpos.2025.1504520/full
16. Artificial Intelligence (AI) and Election Administration | U.S. Election …, accessed June 8, 2025, https://www.eac.gov/AI
17. Political Machines: The Rise of Automated Governance and the …, accessed June 8, 2025, https://theglobalobservatory.org/2025/03/political-machines-the-rise-of-automated-governance-and-the-global-challenge-to-citizens-rights/
18. Voting Rights | American Civil Liberties Union, accessed June 8, 2025, https://www.aclu.org/know-your-rights/voting-rights
19. Your Guide to Federal Voting Rights Law – Department of Justice, accessed June 8, 2025, https://www.justice.gov/crt/media/1229951/dl?inline=
20. Accessible Voting on Election Day – Ohio Secretary of State, accessed June 8, 2025, https://www.ohiosos.gov/elections/voters/voters-with-disabilities/election-day-voting/
21. People-Powered Gen AI – Civic Data Design Lab, accessed June 8, 2025, https://civicdatadesignlab.mit.edu/People-Powered-Gen-AI