As early voting gets underway in the 2026 midterms, a growing number of Americans are experimenting with a new kind of election helper: artificial intelligence chatbots. Rather than calling a friend, consulting a traditional voter guide, or reading through candidate questionnaires, some voters are asking AI tools to research races, fact-check claims and even weigh in on which candidates deserve their support. The trend was highlighted in reporting by NPR's Weekend Edition Sunday.
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 working through his ballot with ChatGPT. Instead of a conversation with a neighbor or relative, Johnson engaged in an hours-long back-and-forth with the OpenAI chatbot, walking through each contest and candidate to determine who aligned best with his values.
Johnson is not a casual user of the technology. He said ChatGPT plays a significant role in his daily life, helping with everything from planning meals to searching for work and handling chores around the house. That familiarity made the chatbot a natural place to turn when he wanted help making sense of a crowded ballot.
His process began with a simple request: he asked the AI tool to list every race he would encounter when he voted. From there, he dug into individual candidates and their positions. The approach reflects how some voters are treating chatbots less as search engines and more as research assistants capable of organizing complex information on demand.
Johnson's experience also hints at the limits of these tools. He noted that in a previous conversation about politics, the chatbot had seemed reluctant to engage, offering only a brief refusal. That kind of caution is common in AI systems, which are often trained to avoid weighing in on elections and sensitive political topics — a constraint that can frustrate users looking for straightforward answers.
Still, the appeal is easy to understand. Midterm ballots can be long and confusing, filled with down-ballot races, local measures and candidates who receive little media coverage. For voters who feel overwhelmed, an AI chatbot offers instant, conversational help at any hour, seemingly tailored to their questions and their location.
Voters are not only using these tools to compare candidates. Some are relying on AI to fact-check political claims, cross-reference statements against news reports and summarize positions that might otherwise require hours of reading. The technology can compress a large amount of information into short, digestible answers, which is precisely what many busy voters say they need.
But the rise of AI-assisted voting raises difficult questions. Chatbots can produce inaccurate or outdated information, and their responses may vary depending on how a question is phrased. Experts worry that voters who place too much trust in a single tool could be misled without realizing it, especially in races where reliable local information is scarce.
There is also the matter of algorithmic bias. AI models are trained on vast amounts of internet text, which can carry the perspectives and blind spots of their sources. Even when a chatbot tries to remain neutral, the way it frames an answer — which sources it cites, which issues it emphasizes — can subtly shape how a voter sees a candidate.
Transparency is another concern. When a voter asks a chatbot for guidance, it is rarely clear where the information came from or how current it is. Unlike a traditional voter guide, which usually lists its authors and sources, an AI answer can appear authoritative while resting on unclear or unverifiable foundations.
The phenomenon is part of a broader shift in how people use AI in daily life. Chatbots have moved from novelty to utility, assisting with writing, planning, shopping and now civic decisions. As the tools become more integrated into routine tasks, it is perhaps inevitable that some users would bring them into the voting booth — or at least into the preparation leading up to it.
Election officials and civil society groups have been watching the trend closely. Some have begun urging voters to treat AI output as a starting point rather than a final answer, and to verify key claims with official sources such as county election offices, nonpartisan voter guides and local news outlets.
For voters like Johnson, though, the appeal of a patient, always-available conversation partner is strong. He described using the chatbot as a way to think through his choices methodically, race by race, rather than relying on gut feeling or a sample ballot alone.
Whether AI becomes a standard part of election preparation remains to be seen. What is already clear is that the 2026 midterms are serving as an early test of how these tools function in real civic life — and of how much responsibility voters are willing to hand over to them.
(0)