Overview
- Google has made the Gemini personal avatar generally available to paid Google AI subscribers and requires users to be at least 18 to create an avatar.
- The feature builds a video likeness from a small set of selfie photos and voice prompts and can produce realistic clips of a user in scenes such as a podcast or at a theme park.
- Google embeds generated videos with its SynthID watermark, limits daily avatar-created videos, and says users can delete an avatar so Google will wipe the underlying selfie and voice data.
- Hands-on testing showed the tool is quick to set up and can produce a convincing false confession when prompted, and a reporter hit a per-user creation cap after three videos.
- The release is part of Google’s wider Gemini Omni multimodal push and raises unanswered questions about enforcement, residual copies, and how regulators and platforms will handle deepfake misuse.