Voters are turning to AI chatbots to compare candidates and check claims ahead of the US midterm elections. The tools can make thinly covered local races easier to research, but an answer that feels persuasive is not necessarily a reliable one.
Watch Desk analysis
What happened
An NPR report published on 5 October describes voters using ChatGPT and Gemini to research candidates, including for down-ballot races. One voter said he asked ChatGPT to map candidates against his own priorities; another used Gemini to compare two candidates in a governor’s race. A volunteer also used Claude to help build a Pennsylvania ballot lookup site.
The report says about half of American adults use AI chatbots, and 42 per cent of chatbot users say they use them to search for information, citing Pew Research Center. Researchers interviewed by NPR warned that agreeable answers can reinforce a user’s existing views, while chatbots may also provide false or misleading information.
Why it matters
Chatbots can gather information and organise comparisons where local reporting is sparse. But a tidy table or confident answer can blur the difference between a candidate’s documented position and a model’s guess. Asking another chatbot may catch some errors, but the researcher NPR spoke to said there was no evidence that this is an effective fact-checking method.
The report offers more useful habits: ask for statements from candidates’ own speeches or official websites, and frame questions without revealing your preferences. One researcher also suggested incognito mode to limit the influence of prior information. These are ways to make a search less leading, not guarantees that the answer is accurate.
Our read
Used carefully, a chatbot can be a research assistant for finding and comparing material, not a substitute for the material itself. Start with what candidates have actually said, check the sources behind each claim, and treat a polished answer with the same healthy suspicion you would give a very confident stranger at a canvassing stall.
NPR’s reporting makes the tension plain: people may gain access to information while leaving more certain of their existing views, rather than better informed. That is a particularly important distinction when the subject is a vote.
What to watch
- Whether chatbot answers link clearly to candidate statements and official sources.
- How voters check claims about local and down-ballot candidates.
- Whether AI companies disclose more about how their models select political information.
Discussion spark: Should voters use chatbots to compare candidates if they check the sources themselves, or is the risk of persuasive, slanted answers too high?
Sources and evidence
- Voters are increasingly using AI chatbots to fact-check and compare candidates – NPR (5 October 2026, 09:00 UTC)
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.