Google is once again allowing the creation of human images with its Gemini AI chatbot, a feature the company had pulled after reports that it was creating historically inaccurate images, such as racially diverse Nazis.
Google announced that an early access version of this feature will be available in English for Gemini Advanced, Business, and Enterprise users “in the coming days.”
This upgrade will be powered by Imagen 3, the latest version of Google’s AI text-to-image generator. Google quietly released Imagen 3 via its AI Test Kitchen earlier this month, and now it’s coming to Gemini in all languages. The improved tool can produce everything from photorealistic landscapes to textured oil paintings with “just a few words” descriptions.
Google discontinued Gemini’s ability to create images of people in February after users were found to have historically created inaccurate images. The upgraded Imagen 3 model comes with built-in security measures and “performs favorably compared to other existing image-making models,” according to a statement from the company.
This is not to be confused with Google’s new Remagine feature, which lets you add AI elements to photos taken on the company’s new Pixel 9 series of phones.
Google Gemini will continue to impose some restrictions
Additionally, Gemini will not allow users to create photo-realistic images of public figures; content that includes minors; or gory, violent, or sexual scenes. “Of course, not every image that Gemini creates will be perfect, but we will continue to listen to feedback from early access Gemini Advanced users as we continue to improve,” said Dave Citron, a senior director at Google. Google plans to expand the ability to create AI-generated personas to more users and languages soon.