Google Puts a Face on Gemini: Live Avatar Arrives for Enterprise Users
Google's new Live Avatar feature gives Gemini Enterprise customers the option of a talking, lip-syncing visual persona for AI agents — raising questions about privacy and the uncanny valley.

Google's Gemini Gets a Face
Google has launched a feature called Live Avatar for customers on Gemini Enterprise, CNET reports. The tool pairs a conversational AI agent with an interactive visual presence, letting businesses attach a face and voice to automated interactions.
According to CNET, the feature arrives days after Google announced Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking.
What It Does
Live Avatar is designed for tasks such as filing an insurance claim through a video chat — an example Google itself used in its announcement, per CNET. Businesses can select from a set of prebuilt avatars to represent the AI agent.
The system supports live video avatars with lip-syncing, and can process audio, screen sharing, and video feeds at the same time, according to the report. It can also run tools and make API calls while maintaining a conversation with a customer.
CNET notes that Live Avatar uses native speech-to-speech to enable more fluid dialogue, including what the report describes as "more natural interruption recovery without dropping conversation context or backend transactions."
Safeguards and Open Questions
Google has outlined some measures to limit misuse. Users choose from curated, prebuilt avatars; creating a custom avatar requires a verification process, according to the company's announcement. All audio and video generated through the feature are watermarked with SynthID, Google's system for marking AI-generated content as verifiable.
Several questions remain unanswered. Google did not immediately respond to CNET's request for comment on when Live Avatar might reach the general public. The company also did not say how it plans to protect information an AI agent receives during video calls — data that could include a user's name, address, and live video feed in scenarios like insurance claims.
Why It Matters
The launch reflects a broader push to make AI agents more personable as businesses adopt them for customer support and other routine interactions. Adding a visual, speaking avatar could make automated service feel more human-like — or, as CNET's reporter put it, "creepy."
For now, access is limited to Gemini Enterprise customers. The feature's trajectory beyond that audience, and the privacy framework around sensitive video-call data, remain unclear.
Reporting attributed to CNET News.
Sources and further reading
See an error? Read our corrections policy or email [email protected].
TECHNOMALIST

