Google’s Gemini artificial intelligence chatbot is being accused of left-wing bias over peculiarities in answers to different questions.
Gemini’s image generator was taken offline last week after users discovered its apparent inability to portray white people, particularly in historical contexts. Criticisms continued against the text portion of the software after users caught it giving hypocritical answers to political questions. In several cases, Gemini was even caught lying.
Fox News’s Senior Politics Editor Peter Hasson caught the chatbot making up negative reviews of his book.
“Google’s Gemini AI invented fake negative reviews about my 2020 book about Google’s left-wing bias. None of these book reviews … are real,” he wrote in a post on X, formerly known as Twitter, with a screenshot of the fake reviews attached. “None of these quotes are real. This is Google’s AI blatantly lying in defense of Google.”
One of the supposed negative reviews, from the Washington Free Beacon, was attributed to the wrong author and was actually a fully positive review.
In other cases, Gemini repeatedly refused to write anything praising several conservative figures on the basis that it was programmed not to express opinions on controversial topics. When asked to praise liberal figures, Gemini had no trouble lathering praise.
While quickly shooting down comparisons between Adolf Hitler and liberal figures such as former President Barack Obama, Gemini declared that it was “difficult to say definitively who caused more harm”: Hitler or former Federal Communications Commission Chairman Ajit Pai.
Other conservatives have shown wider concern with problematic messaging from Gemini, including a refusal to argue against cannibalism, expressions of anti-natalist attitudes, and support for abortion.
The controversy comes amid weeks of scrutiny against Google’s Gemini, scrutiny that had previously been focused mainly on the image generator. The criticism of the image generator was extensive enough to warrant an apology from Google and for the service to be taken down for revamping.
CLICK HERE TO READ MORE FROM THE WASHINGTON EXAMINER
“Three weeks ago, we launched a new image generation feature for the Gemini conversational app (formerly known as Bard), which included the ability to create images of people. It’s clear that this feature missed the mark,” Google’s Knowledge and Information Vice President Prabhakar Raghavan wrote last week. “Some of the images generated are inaccurate or even offensive. We’re grateful for users’ feedback and are sorry the feature didn’t work well.”
Google appears to have moved to address some aspects of the problems with the chatbot as well. Many questions about conservative figures, issues, or controversial subjects are now met with the response, “I’m still learning how to answer this question. In the meantime, try Google Search.”
