As early voting begins across the country for this year's midterm elections, a growing number of Americans are taking an unusual approach to preparing for their trip to the polls: they are consulting artificial intelligence chatbots to help them research candidates and make voting decisions.
The practice marks a notable shift in how voters gather political information, moving beyond traditional sources like news outlets, candidate websites, and conversations with friends and family. Instead, some are turning to large language models (LLMs) such as ChatGPT, the popular tool developed by OpenAI, to walk them through their ballots.
One voter who has embraced this approach is Adam Johnson, a 40-year-old graphic designer from Morgantown, West Virginia, who is studying to become an illustrator. On a September evening, Johnson sat on his porch and spent hours discussing the upcoming midterm races with ChatGPT rather than with another person.
Johnson described the chatbot as a significant presence in his daily life. He relies on it for a range of household and professional tasks, including planning meals and helping with his search for work. When it came to the election, he decided to extend that reliance to his civic duties as well.
His process began by asking the chatbot to compile a list of every race that would appear on his ballot this year. From there, he probed deeper into individual contests, working through the candidates and attempting to determine which ones best aligned with his personal values.
The conversation was not entirely smooth. Johnson recalled that on a previous occasion when he raised political topics, the chatbot seemed reluctant to engage, responding with a simple refusal rather than offering substantive information. That hesitancy reflects the careful guardrails that AI companies have built into their systems around elections and political content.
Despite those limitations, Johnson persisted, and the exchange ultimately stretched over an extended period as he worked through the various races and candidates. His experience illustrates both the appeal and the friction points of using AI as a civic tool.
The trend is not isolated to a single voter. Across the country, people are increasingly using AI chatbots to fact-check claims, compare candidates side by side, and sort through the flood of information that accompanies modern campaigns. For voters who feel overwhelmed by the volume of races and the complexity of local ballots, the technology offers a seemingly convenient shortcut.
Advocates of the practice point to the speed and accessibility that chatbots provide. A voter can ask follow-up questions in real time, request summaries of a candidate's stated positions, or seek clarification on complicated policy issues without navigating multiple websites. For voters who may not have politically engaged friends or family members to consult, the tool can serve as a readily available discussion partner.
Critics, however, warn that the technology carries meaningful risks. Language models can produce inaccurate information, sometimes fabricating details about candidates or races with confidence. Because these systems generate responses based on patterns in their training data rather than verified databases, the information they provide may be outdated, incomplete, or simply wrong.
There is also the question of bias. AI models are trained on vast troves of text drawn from the internet and other sources, and the perspectives embedded in that material can shape the answers they give. Voters who rely on a single chatbot without cross-checking its claims could unknowingly absorb skewed or misleading framings of candidates and issues.
Election officials and technology researchers have urged caution, encouraging voters to treat chatbot output as a starting point for further research rather than a final authority. Verifying information through official sources, such as county election offices and nonpartisan voter guides, remains the most reliable way to confirm what will actually appear on a ballot.
The emergence of AI-assisted voting preparation also raises broader questions about the role of technology in democratic life. If voters increasingly outsource the work of evaluating candidates to automated systems, the reasoning behind their choices may become harder to trace, both for the voters themselves and for those studying electoral behavior.
For now, the phenomenon appears to be expanding alongside the rapid adoption of generative AI tools in other areas of daily life. Just as people now use chatbots for writing, cooking, and job hunting, some are folding election research into the same habit, treating the technology as a general-purpose assistant rather than a specialized political resource.
Whether that habit endures beyond this election cycle remains to be seen. But with voting already underway in the midterms, the sight of a voter spending an evening on his porch interrogating a chatbot about the candidates on his ballot suggests that AI has found yet another foothold in American civic life.
(0)