Key facts
- Sony Group plans to launch an in-car AI agent with an original character by fiscal 2027.
- The AI agent is designed to be a companion, offering personalized conversations and assistance.
- A character named 'Wichef', a rabbit-sheep, has been created to represent the AI.
- The initiative aims to create an emotional connection between users and the AI.
- Sony's Mobility Business Division and Sony Creative Products are collaborating on the project.
Sony Group is preparing to introduce an in-car artificial intelligence agent featuring an original character, with a target launch as early as fiscal year 2027. This initiative aims to establish a presence in the evolving in-vehicle entertainment market, particularly as autonomous driving technology advances.
The AI agent, represented by a character named 'Wichef'—a rabbit-sheep hybrid—is designed to go beyond typical robotic assistants. Sony's Mobility Business Division and Sony Creative Products Inc. (SCP) are collaborating, combining Sony's advanced sensing, sound, and interaction technologies with SCP's expertise in character intellectual property. The core objective is to foster an "emotional connection" and make the AI feel like a companion rather than just a tool.
According to Yusuke Takai, leader of project planning from the Mobility Business Division, the AI will analyze a driver's personality and surroundings to create personalized conversations and suggestions. This approach aims to make the AI "someone you want by your side," learning user preferences and adjusting its tone and word choice over time. The assistant will also be accessible via smartphones for a seamless experience both inside and outside the car.
This development builds on Sony's broader AI research, including projects like Gran Turismo Sophy, which demonstrated advanced AI capabilities in racing simulations. The Gran Turismo Sophy project, developed over five years, showcased how research can translate into products and platforms, exploring reinforcement learning and the importance of understanding the problem domain, such as racing dynamics and etiquette.