Google announced a temporary suspension of the Gemini AI-powered image generation feature after discovering it was providing inaccurate information in historical images.
Users on social media platforms complained that the AI tool was creating images of historical figures with different appearances, such as the founding fathers of the United States, deeming it inaccurate.
Google’s decision to temporarily halt image generation for individuals in Gemini came less than 24 hours after the company apologized for the inaccuracies in some historical images produced by the AI model.
Some Gemini users requested images of historical groups or figures and found that the AI had produced images of non-white individuals in the results.
This led to the spread of conspiracy theories online alleging that Google was avoiding capturing images of people with white skin.
Google stated in a post on “X”: “This feature can create images of a large number of individuals, which is considered useful given people’s use of it worldwide.”
They said: “We acknowledge that the technology provides inaccurate information in some historical images created by AI, and we are working to improve this type of image immediately.”
Google issued a statement announcing a temporary suspension of the feature with a re-release of an enhanced version soon.
Google launched the tool in early February, based on the Gemini AI model, previously known as Bard.
The feature faced challenges around the same time as Google’s attempt to compete with OpenAI backed by Microsoft.
Last week, OpenAI launched “Sura,” a new AI model capable of creating videos based on user-requested texts.
Jack Krawczik, a senior product manager for Gemini products at Google, explained that the company’s capabilities in creating images reflect the global user base of this search giant and that it works diligently and fairly in representation and neutrality.
Krawczik stated in a post on X: We continue to do this for demands with open-ended results. Historical contexts hold many subtle differences, and we are trying to organize them for understanding.
Google launched the Gemini AI model at the end of 2023, and this month renamed its brand from Bard to Gemini, also launching a subscription service for a more powerful version of the AI model.
