openKylin 3.0 Brings AI Agents Into the Linux Desktop Architecture

Summary

China’s community-developed openKylin Linux distribution is positioning AI as part of the operating system’s architecture rather than as a separate chatbot. In her Open Source Summit Europe keynote, executive director Liu Min described the KylinBot framework as a way to manage user intent by coordinating models, data, tools, and devices within defined permissions, while retaining conventional resource management as the foundation. openKylin 3.0 says KylinBot can accept voice and chat requests, retain context, and invoke system capabilities; the Xiao K assistant interface is built on top of the framework. Integrated Kylin-CUA extends agent access to settings and device-management functions, including input devices, displays, networking, Bluetooth, and power controls, while third-party tools such as OpenClaw and WorkBuddy are also supported. The described architecture layers Linux, drivers, and hardware beneath local inference and shared AI runtime services, planning agents, and computer-use automation for graphical applications. Alongside this AI work, the release uses Linux 7.0, optimizes more than 180 core components, improves RISC-V support, expands Rust in system tools, and explores content-based intelligent search. The article says these capabilities were presented as development work rather than independently demonstrated results. openKylin remains a work in progress and is primarily a desktop upstream for the commercial Kylin operating system; smart devices, industrial automation, robotics, high-performance computing, and cloud environments are development directions rather than equally mature platforms. Developers identify agent authorization, permitted actions, governance, and human intervention as continuing security priorities.