AI Chatbots Give Inaccurate Voting Answers
Left 100%
Center 0%
Right 0%
2 left · 0 center · 0 right
What happened
Researchers with the Institute for Strategic Dialogue (ISD) released a report finding that nearly a third of AI chatbot replies to voting questions were inaccurate or outdated. The study tested 15 generic prompts across six AI chatbots. The report was issued as voters have begun using chatbots as a source of information about upcoming elections.
Sourcing check · Mostly on the record · 3 articles
1 of 2 articles cite a named source for the story. It was on the record from the first report.
What these terms mean and how they're counted
- On the record · names a source 1 of 2
- On the record means the article cites a source by name for the story's main claim: a person, or an institution speaking officially, such as a company statement or a named agency's spokesperson. The opposite is an unnamed source. The count is how many of the counted articles cite at least one such source; the names are everyone named across them, each once however the articles spell it. Here: Institute for Strategic Dialogue
- Primary evidence · cites a document or data 1 of 2
- Primary evidence means the article cites a record you could check yourself: a court filing or ruling, a dataset or report, a recording or transcript, an official statement or press release, market data. A quote is not a document. The count is how many articles cite one. Here: study
- Credits another outlet 0 of 2
- The article says the story comes from another outlet's reporting (“first reported by Reuters”). Those articles relay the story; they are not independent confirmation. Here: Each article presents its own reporting.
- Unnamed sources 0 of 2
- The main claim is attributed to people who are not identified: “people familiar with the matter”, “officials said”. The count is how many articles do this.
- How it developed
- When the first report landed and whether it named anyone; when the first named source and the first document appeared. Here: First report Sep 4, 16:57 UTC by The Hill, naming its source. First document 6 hours later, from The Hill (study).
- Not counted 1
- Opinion, analysis and video pieces, and articles too short to assess. Every count above is out of the counted articles. An article republished under a new headline is one article.
This is about who is cited, not whether the story is true. Each source below says what it names. How we count
Omitted — what each side leaves out
Unpacked
Right-leaning coverage does not report that chatbots gave inaccurate or outdated answers to voting questions in nearly a third of tested replies. Left-leaning coverage at The Hill is simply more complete on the tests, saying researchers with the Institute for Strategic Dialogue put generic prompts to six chatbots and found that rate of bad or stale answers. Anyone who read only the Washington Examiner would not know that study of election Q&A existed.
Left-leaning coverage opens on chatbots becoming a source for voters asking about upcoming elections. Right-leaning coverage instead opens by saying Americans "did not show up on any recent Election Day and vote for new technology." The voting-question tests never appear once the piece turns to who should write AI guardrails.
Which chatbot produced the bad voting answers, and would acting on them have cost someone a ballot?
Bottom line
If your AI news came from the Examiner, you missed The Hill's study showing chatbots already failing nearly a third of voting-question replies.
The receipts — 3 facts one side's coverage lacks
- Nearly a third of AI chatbot replies to voting questions were inaccurate or outdated
- Researchers tested 15 different generic prompts
- The ISD report was released Thursday
The Left View
The Hill presented the ISD results as a warning about election information quality, stressing that chatbots are becoming a place voters go for voting questions and that nearly a third of the tested replies “could be inaccurate or outdated.” Coverage framed the story around the reliability of AI-generated answers rather than around industrial policy or geopolitical competition.
The Right View
The Washington Examiner piece provided does not address the ISD voting-questions study. Janet Kelly, CEO of the Alliance for a Better Future, instead argued that “more than three-quarters of Americans simply do not trust AI or its builders,” that “Silicon Valley’s values are often at odds with those of everyday Americans,” and that the live question is whether AI guardrails are “written by Americans, for Americans, or handed to us by foreign regulators and Silicon Valley lawyers.” She cast China as “running hard with no intention of playing fair,” said President Donald Trump brings “the right instincts” on a planned executive order for national security guardrails, and defined success as AI that protects children and the elderly, keeps financial information private, and “creates work rather than erasing it.”
Our Take (balanced)
The left-leaning coverage’s strongest argument is the ISD’s empirical result itself: a controlled set of voting prompts produced a high error rate, which matters because voters are already asking chatbots those questions. The right-leaning source’s strongest argument is a trust-and-control claim backed by Kelly’s assertion that most Americans “do not trust AI or its builders” and that the decisive contest is who writes the rules and whether the United States, not China, sets the terms. The central unresolved tension is whether the story that demands attention is chatbot inaccuracy on voting, which the right-leaning source does not engage, or whether that accuracy problem is secondary to who governs AI and whether American-led development can earn public trust.
3 sources
More stories
The week's bottom lines, in your inbox
One email a week: the five stories that mattered and what they actually mean. Free.