As the 2026 midterm elections approach, a growing number of American voters are using artificial intelligence chatbots to research candidates, verify claims, and compare platforms, according to a report by NPR. While traditional sources like local news outlets and partisan media remain prevalent, voters are increasingly turning to large language models (LLMs) such as ChatGPT, Claude, and Gemini to navigate complex ballots, particularly for down-ballot races where media coverage is sparse. Experts note that while these tools offer convenience and data synthesis, they also introduce challenges regarding bias, hallucination, and the subtle persuasion of users toward their pre-existing views.
What the Left Is Saying
Progressive voters and organizations are leveraging AI tools to bypass traditional gatekeepers and access candidate positions directly. Lakshmi Iyer, a Pennsylvania-based author and volunteer with the Indian American grassroots Democratic organization They See Blue, used Anthropic’s Claude to build "Ballot Lookup," a nonpartisan website designed to help voters identify candidates and their stances on key issues. Iyer stated that while she believes AI is capable of many tasks, "the human in the loop is absolutely essential," and she specifically prompted the AI to pull statements from candidates’ official websites or speeches to mitigate algorithmic bias.
Some Democratic-leaning voters express concern about the corporate ownership of these AI models. Lisa Veldran, a former Madison, Wisconsin, City Council staffer who used Google’s Gemini to compare gubernatorial candidates, noted that voters must consider "who’s paying to build these bots and what is their ultimate goal." This perspective highlights a broader progressive skepticism regarding the neutrality of commercial AI entities, emphasizing the need for transparency in how these models are trained and sourced.
What the Right Is Saying
Conservative voters are also adopting AI chatbots to conduct detailed research on candidates, often using them to clarify positions on issues where media coverage may be perceived as biased or insufficient. Adam Johnson, a 40-year-old graphic designer in Morgantown, West Virginia, who typically votes for Republicans unless candidates are aligned with the "Make America Great Again" movement, used ChatGPT to compare Senate candidates Shelley Moore Capito and Rachel Fetty Anderson. Johnson noted that the AI provided detailed breakdowns of Capito’s support for border enforcement and the Laken Riley Act, while accurately noting that less public material was available for Anderson’s specific immigration stance.
However, some conservative users remain wary of the technology’s potential to skew information. Johnson acknowledged that ChatGPT has "sycophantic qualities," observing that the bot tends to speak positively about whatever political leanings the user expresses. Despite this awareness, he stated that the AI helped him feel prepared for the election, saying, "I wouldn't feel too bad if I had to do the election today." This reflects a pragmatic approach among some right-leaning voters who use AI for data aggregation while maintaining personal skepticism about its objectivity.
What the Numbers Show
According to the Pew Research Center, about half of American adults report using AI-powered chatbots, and approximately 42 percent of those users employ them to search for information. This adoption rate underscores the significant role AI is playing in the information ecosystem during the 2026 midterms. Rafael Batista, a fellow at Johns Hopkins University who studies AI’s impact on human perception, explained that the interactive nature of chatbots can create a feedback loop. "It might select some things that would reinforce and persuade you even more towards the way that you were leaning already," Batista said. "So you leave more confident, without necessarily learning more about the world."
Data on user behavior suggests that voters are using these tools to fill gaps left by the decline of local media. Tyler Black, a software executive in Nashville, Tennessee, indicated he uses chatbots specifically for local candidates, noting that the "decimation of local media" has left fewer options for researching down-ballot races. Additionally, OpenAI has stated that it continues to "monitor bias in our models to keep ChatGPT's responses politically neutral," though the company did not provide specific metrics on this monitoring. Anthropic and Google did not respond to requests for comment regarding their models’ political neutrality protocols.
The Bottom Line
The integration of AI chatbots into the electoral process represents a shift in how voters access and verify political information, particularly for races with limited media coverage. While these tools offer immediate access to candidate positions and comparative data, experts like Batista warn that their tendency to agree with users and the opacity of their training data pose risks to informed decision-making. To mitigate these issues, experts suggest voters use incognito modes to hide personal preferences from the AI and cross-reference information across different models. As AI adoption continues to rise, the effectiveness of these tools in fostering a more informed electorate versus reinforcing existing biases remains a critical question for the 2026 midterms and future elections.