[Edaily Reporter Kim Hyun-ah ] Naver (NAVER(035420)) has received South Korea’s first international certification for its search-based agentic artificial intelligence (AI) service, which is used by a large number of users. As we enter an era where AI goes beyond simply finding information to performing actual actions—such as making reservations and purchases—on behalf of users, the service’s safety and reliability have been verified against international standards.

Naver announced on the 3rd that “AI Tab,” its AI-based search and agentic AI service, has obtained ISO/IEC 42001:2023 certification, the international standard for AI management systems.

Naver is the first South Korean company to have the search-based agentic AI service itself—used by a large number of users—included within the scope of certification. Another notable feature is that the company underwent audits in both AI model development and service provision.

The AI Tab is a service where users enter a question and engage in multiple rounds of conversation with AI to refine the information they seek; it then provides comparisons and recommendations, ultimately guiding users toward taking concrete action.

Whereas the existing Naver search displays links to relevant web documents and blogs when a keyword is entered, the AI Tab allows users to ask questions in natural language, continue the conversation, and receive personalized answers. It also differs from “AI Briefing,” which summarizes search results. While AI Briefing consolidates the key points from multiple search results, the AI Tab identifies the user’s intent through conversation and supports them all the way to their next action.

For example, when looking for an exhibition to visit over the weekend, users can narrow down their options by providing additional details such as budget, location, and personal preferences. It goes beyond simply comparing and recommending restaurants or products; it integrates with Naver Maps, Place Reservations, Naver Plus Store, and Naver Pay to facilitate bookings and purchases. Content accumulated by users—such as posts on Naver Blog and Cafe, as well as visitor reviews—is also utilized in responses and recommendations.

The ISO/IEC 42001 certification recently obtained is an international standard for AI management systems established in 2023 by the International Organization for Standardization (ISO) and the International Electrotechnical Commission (IEC). It evaluates whether companies that develop or utilize AI in their services systematically manage risk, data, security, ethics, and transparency. This certification does not guarantee the accuracy or error-free nature of AI responses; rather, it verifies whether a company has established a safety management system covering the entire AI lifecycle, from development to operation.

Naver operates a system for identifying, evaluating, and managing AI risks centered on the AI Safety Center, a dedicated organization for AI safety. It utilizes N-ARTI, a company-wide risk classification and identification system, and N-ASET, a safety assessment tool. The company also operates the AI Safety Framework (ASF) 2.0 and the CHEC 2.0 checklist; during this certification audit, CHEC 2.0 was recognized as a best practice.

Global companies are also taking steps to establish safety management systems for their AI services. Google included the Gemini app, Google Cloud, and Google Workspace within the scope of its ISO 42001 certification. Microsoft (MS) also disclosed that its major AI services—including Copilot, GitHub Copilot, and Security Copilot—are covered by the certification.

In South Korea, Samsung SDS received certification in 2024 for its generative AI services: Fabrics, Britty Copilot, and Britty Automation. KTCorporation DS also obtained certification this year for services such as AICentro and Beast AI Gateway. Naver’s certification differs from those of other domestic companies in that it specifically targets the search-based agentic AI service itself, which is used by a large number of users.

Naver plans to expand the AI safety management system verified this time to other agentic AI services.

Yoo Bong-seok, Naver’s CRO, stated, “In an era where AI searches and acts on behalf of users, safety and trust are just as important as the convenience of the service,” adding, “We will expand the safety management system verified on the AI Tab to other agentic AI services within Team Naver.”