Microsoft appears to be moving forward with a new portrait experiment inside Copilot titled "Talk to a version of yourself from the future". The feature surfaced alongside several other portrait-related experiments, but this one seems positioned for earlier rollout. It introduces a distinct flow where the user is prompted to take a front-camera photo, which Copilot then uses to generate a personalised avatar. Once created, the system presents a conversational interface where the user can speak with this rendered version of themselves.
The leaked features point to a model that builds a static or semi-animated portrait tied to the user’s likeness, with the primary limitation being that it relies on a single captured image rather than a full body scan or live feed. It also appears to work within Copilot’s existing chat surface rather than a standalone module.
The company behind it, Microsoft, has been expanding Copilot into a broader personal assistant that blends productivity, search, and visual agents. Introducing personalised portraits aligns with the platform’s direction of building more persistent digital counterparts for tasks and decision support.
This feature would likely appear inside Copilot’s experimental section or the visual persona setup screens, where other portrait prototypes have been spotted. It fits into a wider push across the industry to explore human-styled AI constructs that make the assistant feel present and familiar for professionals, creators, and everyday users. The discovery was made through internal interface traces visible to a limited group of testers.
