The feature, which was first posted by Artem Russakovsii on Twitter, is part of Google Workspace Labs. That means only a few people that are part of the test will get to use it.

As Google’s support page explains (via XDA Developers), to use these features, users who are part of the test have to click Apply Visual Effects > Generate a background to open the prompt. You can type in a prompt like “Luxurious living room interior” and even select a style like “Photography,” “Illustration” or “Fantasy” for reference. You can also see other suggestions for generated backgrounds based on the prompt.

In case a user finds generated content unsafe, they can provide feedback to Google about it.

Google previously experimented with AI-powered background generation when the company announced a new feature for Android in May at its Google I/O developer conference. In April, Snap also introduced a generative-AI-based background creation feature for Snapchat+ subscribers.

The search giant has rolled out a good set of features for Google Meet this year, including emoji reactions, pausing streams of individual tiles, support for 1080p streaming and video tile pairing for education users.