Early voting for the midterm elections is already in progress, and a growing number of Americans are experimenting with a novel approach to preparing for their trip to the polls: consulting artificial intelligence chatbots to research their ballots and weigh the candidates.
Rather than relying solely on traditional sources such as voter guides, news coverage, or conversations with friends and relatives, these voters are turning to tools like ChatGPT, the conversational assistant built by OpenAI, to help them sort through the races they will decide.
One example comes from Morgantown, West Virginia, where Adam Johnson, a 40-year-old graphic designer who is studying to become an illustrator, spent a September evening on his porch reviewing his ballot. Instead of talking through the choices with another person, he engaged in a lengthy back-and-forth with ChatGPT, working through each contest and candidate in an effort to determine who best matched his values.
Johnson said the chatbot has become a regular presence in his daily life. He relies on it for tasks such as planning meals, helping with his job search, and handling various chores around his home. That familiarity made it feel natural to bring the tool into his civic decision-making as well.
His process began with a simple request. Johnson asked the chatbot to compile a list of every race that would appear on his ballot this year. From there, he drilled down into individual contests, seeking information about the people running and where they stood on the issues.
The conversation did not always go smoothly, however. Johnson noted that on a previous occasion when he raised political topics, the chatbot appeared reluctant to engage, offering what he described as a simple refusal rather than substantive discussion. That hesitancy reflects the broader caution many AI companies have shown toward election-related content.
The trend Johnson represents appears to be part of a wider shift in how voters prepare for elections. According to reporting from NPR, people are increasingly using AI chatbots not only to compare candidates but also to fact-check claims they encounter during campaign season.
The appeal is easy to understand. Modern chatbots can process enormous amounts of text, respond in plain language, and tailor their answers to a user's specific questions. For voters facing crowded down-ballot races they know little about, that combination can feel like a convenient shortcut.
Yet the practice also raises questions about accuracy and influence. Large language models are known to occasionally produce incorrect or fabricated information, a phenomenon often called hallucination. When those errors involve candidates, voting deadlines, or polling locations, the consequences for voters could be significant.
There is also the matter of how these systems handle political neutrality. Companies behind the leading chatbots have generally tried to avoid appearing to endorse candidates, sometimes declining to answer political questions outright. Users like Johnson have encountered that restraint firsthand, discovering that the same tool willing to help with meal planning may balk at weighing in on a Senate race.
Beyond individual users, the development could reshape how campaigns and civic organizations think about voter outreach. If people are asking chatbots for guidance, the information those systems surface — and the sources they draw from — may become an increasingly important part of the information ecosystem surrounding elections.
The timing of this shift is notable. Early voting is already underway across the country, meaning many voters are making their decisions now rather than waiting until Election Day. Those who turn to AI for help are doing so in real time, as ballots are being cast.
For now, the phenomenon appears to be driven largely by personal curiosity and convenience rather than any organized effort. Voters such as Johnson are experimenting on their own, using tools designed for general conversation to address a specific and consequential task.
Whether the practice spreads more widely remains to be seen, but its emergence reflects how thoroughly AI assistants have woven themselves into everyday routines. A technology first adopted for mundane chores is now being tested on one of the most personal decisions a citizen makes.
The story highlights a broader tension of the AI era: as these systems become more capable and more familiar, people naturally push them into new territory. Elections, with their complexity and high stakes, may prove to be one of the most consequential testing grounds yet.
(0)