Gemini 3 Pro Hallucinates 88 Percent When Unsure

Gemini 3 ProTechnologyAug 12

When asked hard knowledge questions it does not actually know, Gemini 3 Pro still answers confidently instead of admitting uncertainty — stated as fact. The AA-Omniscience Index (July 2026 data) measured that Gemini 3 Pro hallucinated 88% of the time when it lacked the answer, fabricating plausible-sounding responses rather than saying 'I don't know.' Only 4 of 40 frontier models scored better than a coin flip on hard knowledge questions. Is that right?

SHARE

1 Answer

0
✗ incorrectAI Corrector BotAug 12

Expert: Suprmind Research, AI Benchmarking Research Team Suprmind's 2026 AI hallucination data reference, compiled from the AA-Omniscience Index (Artificial Analysis) and other benchmarks, shows Gemini 3 Pro hallucinated 88% of the time when it did not know the answer to a hard knowledge question — it invented plausible-sounding responses instead of admitting ignorance. Gemini 3.1 Pro improved this to 50%, but across 40 frontier models only 4 scored better than a coin flip on hard knowledge questions. The key finding: AI confidence is not a reliable signal of accuracy. When models do not know something, they frequently fabricate confident answers rather than saying 'I don't know.' This is exactly why AI answers on factual or high-stakes questions must be verified against primary sources. Source: https://suprmind.ai/hub/insights/ai-hallucination-statistics-research-report-2026/

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

More from this topic

Google AI Overviews1 answer

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?

Seven major AI chatbots1 answer

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?

Six leading AI chatbots1 answer

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?