Saudi artificial intelligence company HUMAIN and Silicon Valley-based KORA Systems have entered a joint engineering partnership to develop an AI-native Linux operating system for enterprise computers. Announced at the LEAP 2026 technology conference in Riyadh, the proposed platform is designed to interpret a user’s objective and coordinate the applications, AI models and computing resources needed to carry it out.
The companies presented an early version of HUMAIN OS at the conference and are targeting commercial availability in 2027.
Development work will cover the operating-system architecture, Linux integration, hardware support, security, device management, AI runtime and user experience.
KORA is contributing operating-system and edge-computing expertise, while HUMAIN is bringing its AI models, infrastructure, enterprise platforms and agent-based technology to the project.
Tareq Amin, chief executive officer of HUMAIN, said: “AI cannot simply be another application sitting on top of the desktop. Intelligence has to become part of the operating system itself. KORA brings deep operating-system & edge AI expertise, while HUMAIN brings the full AI stack, from models and agents to enterprise platforms and infrastructure. Together, we are building from the operating-system layer upward and rethinking what the personal computer can become when intelligence is part of the architecture from the beginning: from applications to intent, from intent to intelligence, and from intelligence to action.”
From User Instructions to Automated Actions
Conventional desktop systems generally require users to choose an application and work through a series of steps. HUMAIN OS is being designed around a different sequence: a user communicates an objective, the system gathers permitted contextual information, chooses an appropriate model or service and decides where the task should run.
Kirt McMaster, chief executive officer of KORA Systems, said: “Operating systems were built for a world in which the user had to tell the computer every step required to complete a task. This partnership gives us the opportunity to rethink that model. HUMAIN brings the intelligence and enterprise AI capabilities, and KORA brings the operating-system& edge AI capabilities. Together, we are building toward a platform where the user expresses the objective and the system determines how the work gets done.”
A request involving documents, data analysis and presentation software, for example, could potentially be coordinated across several tools without requiring the user to open and manage each program separately. Planned functions include semantic search across authorized information, access to enterprise files, AI-assisted document creation and agents capable of controlling computer functions. These remain development objectives rather than independently tested features.
The operating system is also intended to distribute AI workloads among the computer, an organization’s internal systems and cloud infrastructure. Decisions could depend on privacy requirements, network availability, response time, cost and company policy. Such orchestration would place the operating system in charge of more consequential decisions than a conventional desktop assistant, making identity controls, permission boundaries and auditable actions central to enterprise adoption.
Local AI Forms Part of the Design
HUMAIN and KORA are developing an on-device AI layer optimized for Qualcomm computing platforms. HUMAIN is targeting a model with 26 billion parameters that could run locally on a Qualcomm neural processing unit, or NPU.
Keeping some processing on the computer can reduce dependence on an internet connection and may help organizations retain sensitive information locally. Qualcomm describes its Hexagon NPU as dedicated hardware that accelerates AI workloads across laptops and other devices.
Local execution would not eliminate the need for cloud infrastructure. More demanding models or tasks requiring access to centralized company systems could still be sent to private or public cloud environments. HUMAIN OS is therefore being designed as a management layer spanning local devices, enterprise networks and remote computing capacity rather than as an exclusively offline system.
Linux Provides the Foundation
Linux will underpin the new operating system, giving the developers access to an established open-source ecosystem.
Application compatibility could become one of the project’s most important commercial tests. Businesses typically rely on established productivity, security and management software, and any new operating system must fit into those environments without creating excessive migration costs. The companies said they plan to support enterprise identity, governance, policy enforcement and controlled deployment of AI.
Security will carry particular weight because HUMAIN OS is intended to let AI agents take authorized actions on a user’s behalf. Its ability to balance automation with organizational oversight could play an important role in how businesses assess the platform.
HUMAIN Expands Its AI Portfolio
HUMAIN was launched by Saudi Arabia’s Public Investment Fund in May 2025 to work across data centers, cloud infrastructure, AI models and applications. Its portfolio places the operating-system project within a wider effort to build multiple layers of the AI technology stack rather than operating only as a software developer.
HUMAIN OS extends the company’s effort to place AI at the center of personal computing. The Linux-based platform is being developed around user intent, allowing the system to identify an objective, select the appropriate technology and determine where the task should run.
By developing technology at the operating-system level, HUMAIN is connecting its work in AI models and infrastructure more directly with workplace computing. The project positions HUMAIN OS as the layer through which users could access and coordinate those capabilities during everyday enterprise tasks.
