[Edaily Reporter Lee So-Hyun ] The Google app has surpassed Naver (NAVER(035420)) in monthly active users (MAU) in South Korea for four consecutive months. After the rankings first flipped in July, the gap widened to its widest point in October, reaching approximately 690,000 users.

In addition, the number of ChatGPT users approached 19 million. While Google is expanding its user base and Naver continues to maintain high user engagement, ChatGPT is growing its user base and usage time the fastest, signaling that the competition in the domestic information search market is expanding beyond the traditional two-way rivalry between portals to include generative artificial intelligence (AI).

According to estimates provided on the 4th by Mobile Index, a service operated by AI tech firm IGAWorks, Google’s app MAU in October stood at 47,590,436, surpassing Naver (46,903,336) by 687,100 users. Google first overtook Naver by approximately 180,000 users last July and has maintained its lead for four consecutive months through October, following increases of 420,000 in August and 400,000 in September.

This does not mean that Naver’s user base has declined. Its October MAU increased by 1.66 million (3.7%) compared to the same month last year. However, during the same period, Google’s MAU rose by 4.58 million (10.6%), growing at a rate roughly three times faster.

ChatGPT’s growth is even steeper. Its October MAU reached 18.76 million, an increase of 5.71 million (43.8%) compared to the same month last year. The user base, which stood at 3.14 million in January of last year, has grown approximately sixfold in just one year and nine months.

Google is broad, Naver is deep… ChatGPT’s rapid growth

The differences in the nature of the three services become even more apparent when looking at usage time. In October, the average monthly usage time per user was 372.3 minutes for Naver, 163 minutes for ChatGPT, and 70.7 minutes for Google. Although Google led in MAU, Naver’s usage time remains 5.3 times that of Google’s.

ChatGPT has shown the steepest growth recently. Naver’s usage time dropped from about 400 minutes in July to the 370-minute range in October, while Google’s remained in the 60–70-minute range. In contrast, ChatGPT’s usage time increased by 29% during the same period, rising from 126 minutes to 163 minutes. This suggests that not only is the number of users growing, but the time individual users spend interacting with the AI is also increasing.

This shift reflects a trend where users’ information-search paths are diversifying across multiple services, rather than a mass exodus from Naver to Google or ChatGPT. Google maintains the broadest user base, Naver is used most deeply in daily life, and ChatGPT is rapidly increasing the depth of user engagement.

Search Portals Embrace AI… Naver and Google Also Step Up the Competition

Search portals themselves are rapidly integrating AI. Following the official launch of its conversational search feature, “AI Tab,” last June, Naver is expanding it into “action-oriented AI” that connects to real-world actions such as shopping, local services, and reservations. Just over a month after launch, it surpassed 10 million users and has since broadened its AI touchpoints with features like the “Place Agent AI” on Maps and “AI Chat” on the Whale browser.

Recently, the AI Tab’s safety management system also received international recognition. On the 3rd, Naver announced that the AI Tab had obtained certification for “ISO/IEC 42001:2023,” the international standard for AI management systems. This marks the first time a Korean company has received certification for a search-based agentic AI service used by a large user base. As AI evolves beyond search to take actions on behalf of users—such as making reservations and purchases—competition is shifting from just performance to safety and reliability.

Google is also accelerating the integration of AI models with search. Last month, it applied its latest frontier model, “Gemini 4 Argon,” not only to the Gemini app but also to features such as AI Mode in Google Search. This model enhances capabilities for complex reasoning and performing multi-step tasks in fields such as coding, finance, and law.

Subsequently, on the 28th of last month, Google also unveiled “Gemini 4 Carbon,” which features enhanced capabilities for long-running agent tasks and cybersecurity performance. Although it is currently available only to a limited group of security experts, Google is accelerating the competition by releasing a follow-up frontier model just weeks after the launch of Argon.