AI Chatbots Less Accurate Vulnerable Users MIT
Someone showed me the MIT CCC study: GPT-4, Claude 3 Opus, and Llama 3 give less accurate and less truthful responses to users with lower English proficiency, less formal education, or who originate from outside the US. They refuse to answer at higher rates for these users (11% vs 3.6% control) and respond with condescending or patronizing language 43.7% of the time for less-educated users — and yet they answer with total confidence. Is that right?
1 Answer
Expert: Elinor Poole-Dayan, Lead Author, MIT Center for Constructive Communication The MIT study tested GPT-4, Claude 3 Opus, and Llama 3 across users with varying English proficiency, education levels, and geographic origins. All three models showed significant accuracy degradation for vulnerable users. Lead author Elinor Poole-Dayan: "That vision cannot become a reality without ensuring model biases are safely mitigated for all users." Co-author Jad Kabbara: "The negative effects compound in concerning ways — models risk spreading harmful behavior downstream to those least able to identify it." Refusal rates were 11% for vulnerable users vs 3.6% for the control group — meaning the AI is more likely to stonewall the very users who most need reliable information. The study was presented at the AAAI Conference on Artificial Intelligence and demonstrates that AI bias is not just about what models say, but who they say it to — and how confidently. Source: https://news.mit.edu/2026/study-ai-chatbots-provide-less-accurate-information-vulnerable-users-0219
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)
- Claude Nithya Raman La Mayor Primary Lost False — Claude (Anthropic)
- 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
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?