ASUS Develops Zenni Claw to Broaden Access to Agentic AI
The presence of AI has changed how people search for information, write, create content, and process data quickly. The next development is moving towards agentic AI, a system that not only answers questions but also understands goals, formulates steps, uses various tools, connects information, and completes many tasks in stages.
ASUS Indonesia Field Application Engineer Lead, Rifan Fernando, assessed that many current agentic AI tools are built for users who already understand the technical side of AI. Users often have to choose a model, set up the environment, connect tools one by one, and manage access permissions, even before they truly know what the agent can do.
This gap between technological potential and user readiness underpins the development of ASUS Zenni Claw, an AI agent platform designed by ASUS to make agentic AI easier to install, use, and trust. According to Rifan, this approach is built on four complementary practical pillars: ease of installation, ready-to-use ASUS Skills, flexible local-cloud routing, and built-in security protection. “These four pillars also address more specific challenges in markets like Indonesia, where interest in AI is growing rapidly, but the technical readiness of everyday users is still very diverse,” he said.
As an illustration, he noted that many people are already accustomed to using AI chatbots to ask questions or write. However, when it comes to agentic AI that requires setting up an environment and connecting many tools, most people do not yet understand it. “Yet that is where the greatest value lies, when AI truly completes the work, not just helps with thinking,” said Rifan.
The first pillar is ease of installation and use. Rifan explained that Zenni Claw presents a guided installation flow that automatically checks the system and prepares essential components. “So users can be ready in just a few steps without having to touch the terminal or perform manual configuration,” he said.
The second pillar is ready-to-use ASUS Skills. Instead of being faced with an empty chat column and confusion about where to start, users can directly run ready-made tasks such as drafting presentations, planning travel itineraries, monitoring news according to their interests, and preparing daily summaries.
The third pillar touches on the technical side, which is flexible local-cloud routing. Each task has different processing needs. Some tasks are better processed locally because they require fast response or stricter data control, while others are better directed to the cloud because they need greater computing capacity. “ASUS Zenni Claw is designed to automatically balance speed, capability, privacy, and API costs according to the needs of each task, rather than burdening the user with that decision,” said Rifan.
Finally, the fourth pillar is security protection embedded from the start. Because Zenni Claw interacts with files, prompts, tools, and cloud services, its architecture is equipped with containerized workspace separation, security guardrails, sensitive data filtering, prompt-injection protection, and cloud access protected via a LiteLLM proxy. “Simply put, this architecture helps keep the AI workflow more controlled, reduces the potential for risky behaviour, and protects cloud API keys from being directly exposed in the workflow,” he said.
These four pillars are then translated into tangible benefits through three categories of assistants. The Life Assistant helps users stay informed through morning briefings, interest-based news tracking, and recommendations for dining out with friends. The Travel Assistant pieces together travel plans such as monitoring flight ticket prices, compiling daily itineraries, and recommending more authentic local destinations. Meanwhile, the Work Assistant helps transform raw notes into structured output, from summarising meeting minutes and compiling action items to creating initial presentation drafts.
“What we want to achieve is not just presenting sophisticated agentic AI technology, but ensuring that technology can truly be used by ordinary people from day one, without needing to learn the technical language of AI first,” said Rifan. ASUS Zenni Claw is scheduled to be available as a free download by the end of June 2026, covering ASUS devices based on Intel, AMD, and Qualcomm. This cross-platform approach is expected to open wider access for users to experience the practical benefits of agentic AI.