Google’s AI search tool failed to identify Joe Biden as the winner of the 2020 election. The company blamed a technical glitch, but the incident raises questions about how AI handles established facts.

Joe Biden won the 2020 presidential election. But on Wednesday, October 28, Google’s AI-powered search tool struggled to acknowledge that established fact.

When users asked Google AI Mode who won the election, the system directed them to Federal Election Commission records rather than naming Biden. Yet the same tool correctly identified Donald Trump as the winner of the 2016 and 2024 elections, drawing criticism over the inconsistent responses.

Google says it was a technical glitch.

The company explained that its election verification system encountered a technical gap affecting 2020 election queries and certain 2022 Senate races. Google’s separate Gemini chatbot answered correctly, and the AI Mode issue was reportedly resolved by Thursday.

The timing was particularly sensitive. With the November 3 midterm elections approaching, technology companies are facing pressure to prevent AI-generated election misinformation. Four congressional Democrats recently warned major AI developers about inaccurate chatbot responses involving voting procedures and election information.

The bigger picture: Google’s AI didn’t incorrectly declare another candidate the winner. Instead, it sometimes avoided answering a question whose outcome has been certified for nearly six years.

That matters because AI-powered search is increasingly becoming an entry point for information. When technology treats established facts as unresolved, it risks creating uncertainty where none should exist.

Google attributes the incident to a technical failure, and there is no established evidence of deliberate political interference. But the episode illustrates a growing challenge for AI companies: protecting against misinformation without making verified facts seem debatable.