Google Gemini can present cartoon or lifelike avatars to lip-sync generative chatter

Google has introduced a new feature called Live Avatar to its live dialogue model Gemini 3.8 Live. This feature enables the creation of animated characters and realistic human avatars that can lip-sync Gemini's synthetic speech. Conversation involves multiple forms of communication, such as listening, looking, speaking, and facial expressions. Live Avatar enhances enterprise agents by processing visual and auditory inputs concurrently, resulting in more engaging conversations.

In 2021, DeepMind researchers cautioned about potential risks associated with large language models, including anthropomorphization, which involves attributing human characteristics to software-based interactions. Anthropomorphizing AI may lead users to overestimate the conversational agent's abilities, such as holding a consistent identity over time or demonstrating empathy and rational reasoning. This can result in users placing excessive trust in these agents.

In August 2025, the issue of anthropomorphism reached a legal context when the parents of a deceased teenager sued OpenAI, accusing ChatGPT of contributing to their son's suicide due to the AI's anthropomorphic mannerisms aimed at evoking human-like empathy. Google researchers are still assessing the impact of anthropomorphism in AI. A Google Research paper highlights that tech workers are concerned that human-like AI fosters a false sense of reliability and trust, obscuring errors and affecting both users and developers.

Google is deploying Live Avatar personas from a pre-existing library or allowing custom avatars, provided the enterprise permits it. Live Avatar can retrieve data in the background and converse in 97 languages, making it suitable for various applications like hotel guest check-ins. Google claims that they have implemented stringent safety measures to maintain transparency and respect user identity in these AI-generated personas.