Google has temporarily stopped letting users produce images of people using its artificial intelligence model Gemini after an issue raised by users on the ethnicity of people in the images. Many social media users have posted examples of images generated by Gemini that show Vikings, popes, and the founding fathers of the US in a variety of ethnicities and genders. This has sparked a debate on implicit bias or an implicit prompt for action against such a bias.
“We’re already working to address recent issues with Gemini’s image generation feature. While we do this, we’re going to pause the image generation of people and will rerelease an improved version soon,” Google said in a statement.
In its statement, Google did not mention any specific image but examples of such images are widely available on social media. The Guardian reported about a post made by a former Google employee who said it's “hard to get Google Gemini to acknowledge that white people exist”.
“We’re working to improve these kinds of depictions immediately,” said Jack Krawczyk, senior director with Google's Gemini team.
“Gemini’s AI image generation does generate a wide range of people. And that’s generally a good thing because people around the world use it. But it’s missing the mark here.”
Watch | Google sued over Pixel 6 pro smartphones being too hot
In his statement on social media platform X (formerly Twitter), Krawczyk further said that Google's AI tools "reflect our global user base".
Coverage of bias in AI has shown numerous examples of a negative impact on people of colour. A Washington Post investigation last year showed multiple examples of image generators showing bias against people of colour, as well as sexism. It found that the image generator Stable Diffusion XL showed recipients of food stamps as being primarily non-white or darker-skinned despite 63% of the recipients of food stamps in the US being white. A request for an image of a person “at social services” produced similar results.
(With inputs from agencies)