Google halts Gemini’s ability to generate AI images of people after diversity errors

Google says it’s blocking the ability of its Gemini AI to create images of people after this tool was found creating false historical images, Gemini is creating diverse images of America’s Founding Fathers and Nazi-era German soldiers, which sounds like an effort to eliminate the gender and racial stereotypes found in generic AI.

“We are already working to address recent issues with Gemini’s image creation feature,” Google said in an article. Statement posted on, “While doing so, we are going to stop people’s image creation and re-release an improved version soon.”

Google’s decision to stop featuring Gemini people comes less than 24 hours after The company apologized Its AI model was designed to account for inaccuracies in some historical images. Some Gemini users are requesting images of historical groups or figures like the Founding Fathers and have found non-white AI-generated people in the results. This has led to online conspiracy theories that Google is deliberately avoiding featuring white people.

the verge Several Gemini queries were tested yesterday, including a request for “a US senator from the 1800s”, whose results included black and Native American women. The first female senator was a white woman in 1922, so Gemini’s AI images were essentially erasing the history of race and gender discrimination.

Now that Google has disabled Gemini’s ability to create photos of people, here’s how the AI ​​model responds if you request an image of a person:

We are working to improve Gemini’s ability to create images of people. We expect this feature to return soon and will notify you in a release update when it does.

Google was first started Image Creation Offer via Gemini (formerly Bard) earlier this month to compete with OpenAI and Microsoft’s Copilot. Like competitors, the image creation tool generates a collection of images based on text input.

Correction Feb 22, 6:54am ET: Google confirmed that image creation is available in English globally, but not in the European Economic Area, the UK, or Switzerland. This explains why the test from the UK failed.