Google has paused its artificial intelligence image generator’s ability to create pictures of people to repair its representation of race after it stoked controversy by injecting minorities inappropriately into images.
Google’s Gemini chatbot, released to the public last week, drew widespread criticism for misrepresenting historical figures as different races. For example, Gemini presented white historical figures as black, Native American, or Asian, including the Founding Fathers and the pope. Several conservative users noticed the inaccuracies and blasted the search engine, alleging that it was evidence of a “woke” AI model.
“We’re already working to address recent issues with Gemini’s image generation feature,” Google said in an X post. “While we do this, we’re going to pause the image generation of people and will re-release an improved version soon.”
Jack Krawczyk, the head of product for Google’s AI division, apologized on Wednesday for the inaccuracy, saying that “image generation capabilities reflect our global user base, and we take representation and bias seriously.”
The images attracted ire from several online users. “It’s embarrassingly hard to get Google Gemini to acknowledge that white people exist,” computer scientist Debarghya Das posted.
It is not the first time Google has been accused of stumbling over diversity matters. The company had to apologize almost a decade ago for its photo app labeling an image of a black couple as “gorillas.”
Google Gemini, formerly known as Google Bard, was launched in March 2023 in limited release as a chatbot powered by Google’s large language model. The product was renamed in February as “Gemini” to reflect the “advanced tech at its core” after multiple upgrades to the model, according to Google CEO Sundar Pichai.
