Google Lifts Ban On Gemini AI Generating Images Of People

Google has announced it will soon lift its ban on creating images of people with its Gemini AI. The text to image generation feature fell under heavy criticism after launch when people began noticing Gemini was not keeping things historical accurate.

Shortly after launch, Google decided to pull the ability for its Gemini AI to generate images of people. Some examples shared on social media included renderings of 1943 Nazi Germans being of Asian descent, and US founding fathers being African American. The tech giant quickly responded that it was already working to address the issues, but would be pausing the image generation of people while doing so. Now, it seems the company is comfortable enough to re-release the ability to generate images of humans, and will do so in the coming days.

Perhaps the biggest change to the feature is the addition of Imagen 3 to Gemini. Google remarked in a press release, “Imagen 3 brings advanced image generation capabilities that come with built-in safeguards and adhere to our product design principles.”

It should also be noted that the ability to generate images of humans will be locked behind a paywall at first. Users will be required to have a Gemini Advanced, Business, or Enterprise plan, and will only be available in English to start. As for when the feature will make its way back to Gemini, the company only stated, “over the coming days.”

Another aspect of the feature making a return users should be aware of is that it won’t support certain renderings. Google Gemini’s Senior Director of Product Management David Citron explained, “We don’t support the generation of photorealistic, identifiable individuals, depictions of minors or excessively gory, violent or sexual scenes.”

Citron added Gemini is not perfect, and the company will continue to closely monitor user feedback to improve it. He also remarked that while Google Gemini’s new people-focused AI-generated feature will only support English in the beginning, the company has plans to add more languages soon.

Anyone who would like to give Gemini’s ability to generate images of people, and is not already subscribed to Gemini Advanced, can visit the Gemini Advanced webpage. There is a free month trial available for anyone who has never had a subscription before.