Google rolls out Gemini Avatar, letting users create AI video clones of themselves

Gemini Avatar Tool Rolls Out, Letting Users Create AI Video Clones

Google is now widely rolling out its Gemini Avatar feature, a tool that lets paid subscribers create realistic AI-generated videos of themselves using their own face and voice. The feature, powered by the company’s new Gemini Omni model, was first announced at Google I/O 2026 on May 19 and began reaching a broader audience this week.thenextweb

How It Works

The setup process takes just a few minutes. Users navigate to Settings > Avatar in the Gemini app and follow a guided enrollment that uses the phone’s front-facing camera. The system asks users to look into the camera, move their head from side to side, and read a series of random numbers aloud so Gemini can map their facial structure and capture their voice. Once complete, the digital likeness is bound to the user’s account and can be summoned in any Gemini chat by typing @me or the user’s handle.androidauthority

The results are strikingly lifelike. “My avatar’s resemblance is honestly unsettling because it’s so believable,” Android Authority’s reviewer wrote. “The facial movements and tone of voice are extremely realistic and could easily fool someone who doesn’t know me very well.”androidauthority

Deepfake Guardrails

Google has built in several safeguards to address the obvious misuse potential. Users must be at least 18 years old, and the account owner must be physically present during the setup, speaking the prompted numbers live while on camera — a measure designed to prevent anyone from creating an avatar of someone else using uploaded photos or clips. Every generated video carries an invisible SynthID watermark, Google DeepMind’s provenance tool, which can be verified through the Gemini app, Chrome, or Google Search.deepmind

The company is also withholding general-purpose audio and speech editing inside the Omni model for now. “We are still working to test this and better understand how we can bring this capability to users responsibly,” Koray Kavukcuoglu, CTO of Google DeepMind, wrote in a blog post accompanying the I/O announcement.thenextweb

Broader Context

The Avatar feature is currently available to Google AI Plus, Pro, and Ultra subscribers in the United States and select regions, though it remains blocked in the European Economic Area, Switzerland, and the United Kingdom. It arrives as concerns about AI-generated deepfakes continue to mount; a Fortune report in late 2025 warned that synthetic media had reached a level of realism capable of “reliably fooling nonexpert viewers” in most everyday scenarios. Google itself rolled out a separate deepfake call-detection feature for Android devices this week, underscoring the dual-edged nature of the technology it is now putting in consumers’ hands.google