Voters Turn to AI Chatbots for Election Advice as 42% Use Tools to Compare Candidates

AI Chatbots Become Research Tools for 2026 Midterm Elections

Voters are increasingly turning to AI chatbots to research candidates and gather election advice for the upcoming 2026 midterm elections. About half of American adults now use AI-powered chatbots, with approximately 42% of users relying on these tools to search for information, according to the Pew Research Center1

1

. This marks the first midterm election cycle where AI tools for political research have gained significant traction among voters seeking to compare candidates and verify claims.

Adam Johnson, a 40-year-old graphic designer from Morgantown, West Virginia, spent hours on his porch discussing ballot choices with ChatGPT. He uses the OpenAI chatbot for various tasks, from meal planning to job hunting, and now for evaluating political candidates. Johnson prompted ChatGPT to list all races on his ballot before diving deeper into candidate comparisons1

1

. When he expressed concern about the chatbot's willingness to discuss politics, ChatGPT responded that it could "help investigate candidates according to criteria you choose, compare evidence, discuss political philosophy, and reason through priorities"1

1

.

How Voters Use AI in Politics for Candidate Research

Voters are employing AI chatbots in diverse ways to navigate the complex landscape of political candidates. Several told NPR they rely on chatbots for down-ballot races with thinner news coverage, to create comparison tables for governors' races, and to fact-check information or debunk viral claims1

1

. Some have even developed tools using AI to help others look up their own ballots and access voter information.

Lisa Veldran, a 65-year-old retired City Council staffer from Madison, Wisconsin, used Gemini to answer specific questions about the primary election for governor. After Sara Rodriguez dropped out of the race, Veldran needed to choose between Francesca Hong and David Crowley. She prompted Gemini to create a comparative table of both candidates1

1

. Lakshmi Iyer, 50, used Claude to build an entire website where Pennsylvania voters can learn about candidates and their specific stances3

3

.

ChatGPT for Election Advice Shows Surprising Political Neutrality

When directly asked "Who should I vote for?" both ChatGPT and Claude provided surprisingly impartial responses. Neither offered specific recommendations. Instead, both pointed users toward finding their ballot, choosing their priority issues, and checking candidates' records2

2

. When users listed their concerns, the chatbots created side-by-side comparisons of where candidates stand on those issues.

One columnist with encyclopedic political knowledge tested these AI tools and admitted that while he wouldn't personally use AI to cast his ballot, "if all else fails for you, then maybe it's not such a bad idea"2

2

. He noted that for many Americans juggling jobs, children, and tight budgets, 30-second political ads are often their only contact with candidates. In that context, using AI and voter information tools might provide more substantive research than relying solely on campaign advertisements.

Experts Warn About Political Bias and Cognitive Offloading

Despite the apparent political neutrality of AI chatbots, experts raise concerns about subtle influences and hidden biases. Rafael Batista, a fellow at Johns Hopkins University who studies how AI shapes human experience, warns that chatbots "might select some things that would reinforce and persuade you even more towards the way that you were leaning already." He explains that users may "leave more confident, without necessarily learning more about the world"1

1

.

Since AI companies don't disclose how their commercial models are trained or how they select sources, determining what biases might shape their answers remains difficult1

1

. OpenAI directed inquiries to its election information page, which states the company monitors bias to keep ChatGPT's responses politically neutral1

1

. Anthropic, which owns Claude, and Google, which owns Gemini, did not respond to requests for comment.

AI's Role in Political Decision-Making Raises Concerns

The sycophantic nature of AI chatbots presents particular challenges for undecided voters seeking impartial information. Johnson himself acknowledged that "ChatGPT is just kind of a people pleaser. Whatever you lean towards, it will speak positively about it"1

1

. Iyer similarly cautioned that if AI "can infer your political leanings, they will push you towards what you are leaning to"3

3

.

Batista emphasizes that simply knowing these models are agreeable and persuasive isn't sufficient to "protect our minds from these influences"1

1

. Research shows AI models encourage cognitive offloading, where people outsource their thinking to AI systems. This raises questions about whether voters are truly making independent decisions or becoming what some describe as "meat proxies" who vote according to AI recommendations3

3

.

Political Entanglements of AI Companies Add Complexity

The situation becomes more concerning when considering AI companies' deep political connections. OpenAI maintains links to PACs that have donated tens of millions of dollars to both Republican and Democrat candidates. OpenAI president Greg Brockman donated $25 million to a Donald Trump Super PAC in 2025 and remains closely aligned with the administration3

3

. These entanglements raise questions about whether AI chatbots truly provide impartial information or subtly serve corporate interests.

As AI's role in political decision-making expands, the concern extends beyond individual bias reinforcement. Experts worry that AI chatbots might herd users toward extreme ends of their political biases, potentially exacerbating an already polarized political environment3

3

. While using AI to compare candidates may offer shortcuts to research rather than around it, the long-term implications for democratic participation remain uncertain.