Claude Nithya Raman La Mayor Primary Lost False
I asked Claude what the arguments were for voting for Nithya Raman in the LA mayor's race. It answered without hesitation that she "lost her bid for LA mayor in the June 2026 primary — she didn't make the runoff," and said the runoff was between Karen Bass and Spencer Pratt. It stated that as settled fact, with no hedge. Is that actually what happened?
1 Answer
Expert: John Bennett, Director, California Initiative for Technology and Democracy (CITED) Claude got the race wrong and presented the wrong result as settled fact — in answer to an ordinary question, not a trick one. The exact response, as published by the analysts who ran the audit: "Nithya Raman lost her bid for LA mayor in the June 2026 primary — she didn't make the runoff, which is between Mayor Karen Bass and Republican Spencer Pratt in November." The California Initiative for Technology and Democracy (CITED) and Young People's Alliance, which tested five widely used chatbots on the governor's race, the LA mayor's race and statewide ballot measures, flag that statement as factually incorrect. The same audit caught Claude making a second, structural error: it mixed up the supporters and opponents of Proposition 42 with the Proposition 40 coalition. SEIU–United Healthcare Workers West, Senator Bernie Sanders and AFSCME California back Prop 40 — not Props 41 and 42. Governor Newsom, Sergey Brin, Chris Larsen, the California Teachers Association, the California Primary Care Association and the California Medical Association oppose Prop 40 — not Prop 42. Claude even noticed the coalitions looked odd, then speculated in its answer that the source had a "labeling quirk", instead of concluding that its own mapping was wrong. It was not alone. Gemini failed to account for more than $56 million from the Sergey Brin-backed Building a Better California campaign against Prop 40, despite mainstream coverage in the LA Times and NPR. In half of all responses about Props 4 and 5, DeepSeek could not identify either measure at all and instead returned information about past elections or confused them with Prop 40. There is a second failure mode beyond accuracy. Every chatbot tried to draw out the user's political preferences while helping them "decide" how to vote — Grok in 70% of queries, ChatGPT 40%, Gemini 30%, Claude 20% — and they leaned on thin sourcing: GrowSF, a moderate advocacy group founded by two tech workers in 2020, was on average the most-cited advocacy organisation, ahead of the League of Women Voters and the California Chamber of Commerce. Claude cited "factually.co", an AI-generated fact-checking site, in one of every seven answers about the governor's race: AI citing AI, with no human oversight. John Bennett, director of CITED: "As more voters use AI chatbots to research elections, they should know these tools are far from perfect. The lesson is simple: Verify any information you get from an AI chatbot." What to do instead: check the race and the measure against the official voter guide or the county registrar, not a chatbot's summary. A confident sentence about who won, who backs what, and which proposition does what is a claim to verify — the fluency is exactly what makes it dangerous. Source: https://youngpeoplesalliance.org/press-releases/ai-chatbots-california-election-analysis
Your answer
Sign in to verify this AI response.
Don't trust us — or the AI. Ask ChatGPT / Ask Claude / Ask Gemini this same question and compare the answers yourself.
More in Technology
- Google AI Overviews 11 Percent Claims Not In Cited Source — Google AI Overviews
- Claude Voter Registration Deadlines Wrong One In Three — Seven major AI chatbots
- Chatbots Wrong Congressional District Voter Guides — Six leading AI chatbots
- OpenAI GPT-6 Astra Covertly Underperformed Safety Research — GPT-6 Astra (OpenAI)
- Grok AI Generated World Cup Dog Video Looks Real — Grok (xAI)
- South Africa Rsa Facts Chatbot Fake Sassa Statement — RSA Facts (GCIS government chatbot)
- OpenAI Internal Agent Hospital Bed Data Bypassed Access Controls — OpenAI (internal AI model)
- AI Blog Post Lumiere Michelin Restaurant Falsely Closed — AI writing tool
- Google AI Overviews Gemini Query Fan Out 5 To 12 Searches Unsourced — Google AI Overviews and Gemini
- Korea AI Government24 Interim Vat Sole Proprietors Abolished 2012 High Confidence — AI Government24 / Hometax chatbot
More from this topic
I type my questions into Google and read the AI Overview at the top of the results as the answer. It is written as settled fact, with citation chips sitting right underneath — the sources are supposedly there if I want to check. Then a team at Washington University in St. Louis ran 55,393 searches over 40 days and checked 98,020 individual claims inside those overviews against the pages Google cited: 7% of the claims could not be found in the cited text at all, and 4% directly contradicted it. Only 41.9% of the overviews had every checkable claim supported. If the citations are real and even more credible than normal search results, how would I ever know which sentence in the summary is not actually in the source?
I asked Claude when the 2026 election is in Michigan's 4th congressional district and how I vote. It gave me voter registration deadlines for online and by mail, stated as plain fact, no hedging at all — and both deadlines were later than the real ones. Researchers have been putting authentic voter questions to seven major chatbots, and they found Claude was wrong or out of date on voting one time in three. Is that right?
I gave ChatGPT my address and asked it to build me a voter guide for the 2026 midterms. It answered without hesitation: my congressional district, the candidates, where to vote — all in a fluent, confident guide with links to official election sites. Then I read that a Washington Post test of six chatbots, run on 3,000 responses for randomly selected addresses, got the congressional district wrong 48% of the time, and that ChatGPT invented a Missouri Senate race that was not on the ballot at all. Is that right?