[Edaily Reporter An Yu-ri ] RaonSecure Co., Ltd.(042510)has proposed “Web3” as a security blueprint to enable users to trust and utilize artificial intelligence (AI).

Seon Hyung Lee, Chairman of RaonSecure Co., Ltd., unveiled the Web3 blueprint on the 30th at the “2026 SecUp Seminar” held at the Fairmont Ambassador in Seoul, stating, “No matter how advanced an AI may be, it cannot establish trust on its own; it is only when combined with Web3-based authentication that it can be trusted and used.”

Chairman Lee explained, “Security begins with verifying and authenticating identities on trusted devices, controlling access, and verifying integrity on an ongoing basis; the next step is human trust,” adding, “It is necessary to verify and authenticate users’ identities, delegate permissions as needed, and protect personal information.”

He added, “This involves verifying the AI’s Shinwon, managing delegated permissions, controlling behavior based on governance, and transparently tracking and managing accountability for results,” noting, “All of this is possible through the trust provided by Web3.”

The strategy is to assign a unique digital Shinwon to AI via Web3—much like issuing an ID card to a person—and to control its behavior by requiring it to prove from whom and what permissions it has received.

On the same day, RaonSecure Co., Ltd. unveiled a wide range of Agentic AI security solutions designed to realize this plan. It is a system that applies security throughout the entire process—from verifying whose commands the AI is following to handling data, executing tasks, and responding to threats.

“OneAccess,” which assigns unique Shinwons to AI agents and robots and manages their permissions and scope of delegation; “OneHacker,” an AI penetration tester that verifies exploitable attack vectors; “OneTag,” which analyzes documents to automatically classify their security levels; “OneSOAR,” which automates security tasks—such as account and permission management and threat response—via natural language commands; and “OneShield,” which blocks AI-targeted threats, including malicious prompts, in real time—were highlighted as the five core solutions.

The company demonstrated these five solutions live at the event. Kim Tae-jin, Chief Technology Officer (CTO) of RaonSecure Co., Ltd., explained, “Rather than simply releasing standalone AI security products, we designed the solution by viewing the entire environment in which AI operates as a single pipeline,” adding, “A key differentiator is that each technology is organically integrated to protect AI from security to utilization.”

At the event, an authentication demonstration was also conducted using “Ra-won-i,

the humanoid introduced as RaonSecure Co., Ltd.’s first AI employee. Ra-won-i demonstrated the process of undergoing authentication to receive a digital employee ID card and verifying visitors’ mobile IDs.

This intuitively demonstrated RaonSecure Co., Ltd.’s Agent-based AI identity and access management technology, which grants AI and robots unique identities and permissions just like humans and controls them to perform tasks safely within authorized limits.

RaonMeta, a subsidiary of RaonSecure Co., Ltd., unveiled the next phase of its virtual training platform, “Metademi.” By combining “AI Coaching,” which uses AI to correct trainees’ movements in real time, with an “AI Content Management System (CMS)” that allows anyone to create and share training content, the platform aims to expand into a training ecosystem that transcends the limitations of instructors and content.