Google has announced an update to its Gemini platform, integrating the latest Imagen 3 image generator and resuming the generation of images that include people. Earlier this year, Google paused the depiction of people due to concerns over generating historically inaccurate images. With the integration of Imagen 3, Google is reintroducing this capability but with enhanced safety protocols.

The updated image generator will be available on Gemini in the coming days, and Google is cautiously bringing back the ability to create images with people under strict guidelines. These include avoiding photorealistic depictions, excluding children or identifiable individuals, and preventing the creation of excessively violent or inappropriate content. Initially, this feature will be limited to English-language prompts and available only to Gemini Advanced, Business, and Enterprise users.

Additionally, Google is expanding support for the "Gems" feature, which allows users to create domain-specific versions of Gemini. This feature, previewed at Google I/O, is now accessible in 150 languages for the same group of paid users.