Google Gemini can now use animated or realistic avatars to talk to users

Google has expanded its Gemini 3.8 Live dialogue model by introducing a new capability known as "Live Avatar." This update allows the system to produce either animated cartoon characters or realistic human-like figures that synchronize their lip movements with machine-generated speech. According to Google DeepMind research scientist Shuo-yiin Chang and software engineer CJ Zheng, the technology processes audio and visual data at the same time to create a more immersive, multimodal communication experience for enterprise agents.

This development introduces significant considerations for corporate adoption. While the feature aims to make digital interactions more natural, it heightens the risk of anthropomorphism, where users might mistakenly attribute human intelligence or capabilities to the software. Consequently, organizations implementing these tools should conduct thorough assessments regarding the AI's actual functional limits, data privacy protocols, and overall safety to prevent users from overestimating the competence of the conversational agents.