Microsoft Copilot’s New AI Faces Bring Conversations to Life
▼ Summary
– Microsoft has introduced an experimental “Portraits” feature for Copilot, adding animated human avatars that display natural expressions during voice conversations.
– The feature is currently available to users in the US, UK, and Canada, allowing them to select from 40 stylized avatars and pair them with a voice for more natural interactions.
– Microsoft AI CEO Mustafa Suleyman stated that Portraits was developed in response to user feedback expressing a preference for talking to a face during voice conversations.
– This experiment uses Microsoft Research’s VASA-1 technology to generate real-time facial animations and lip-sync from a single image without complex 3D modeling.
– Microsoft is rolling out Portraits cautiously with age restrictions, time limits, and clear AI indicators due to safety concerns, following similar avatar features from competitors.
Microsoft is introducing animated faces for its Copilot AI assistant, aiming to make voice conversations feel more human and engaging. This experimental Portraits feature, currently accessible to users in the United States, Canada, and the United Kingdom, offers forty stylized human avatars that display natural expressions during real-time spoken interactions.
According to Microsoft AI CEO Mustafa Suleyman, the development of Portraits was driven by user feedback indicating that many people prefer seeing a face when using voice features. By allowing individuals to select both a portrait and a voice, Microsoft hopes to create a more comfortable and intuitive conversational experience compared to standard text exchanges.
This initiative represents Microsoft’s ongoing effort to enhance Copilot’s appeal as a conversational partner. Earlier this year, the company previewed a related feature called “Copilot Appearances,” but Portraits differs by using human-like avatars rather than animated cartoon figures.
Eligible Copilot users can now choose from a range of AI-generated portraits that synchronize facial animations with spoken responses. Microsoft confirmed that the technology behind this experiment is VASA-1, an advanced framework from Microsoft Research that produces realistic facial expressions, head motions, and lip-syncing in real time—all from a single static image, eliminating the need for complicated 3D modeling.
The launch of Portraits comes as other AI providers, such as X with its Grok chatbot, roll out their own 3D avatars—some even offering anime-inspired companions with adult-oriented modes. Microsoft appears to be proceeding cautiously with its own rollout, possibly to avoid the types of harmful user interactions that have prompted scrutiny for platforms like Character.AI.
In a blog post, Microsoft emphasized that each portrait is intentionally stylized rather than photorealistic, providing visual diversity while maintaining the reliability and security users expect from Copilot. Access to the Portraits experiment is currently limited to a small group of adult users aged 18 and above, with built-in time restrictions per day and per session. The company also confirmed that clear indicators are displayed to remind individuals they are interacting with an artificial intelligence system.
(Source: The Verge)
