Naver ends consumer chatbots, doubles down on AI for shopping and search

Naver, South Korea’s dominant internet portal operator, is shifting its AI strategy away from consumer-facing chatbot services toward more profitable AI applications, including shopping assistants. The company said it will terminate its conversational AI service Clova X and its generative AI search product Cue: on April 9, signaling a broader pivot from public chatbots to commercially viable AI tools.

Clova X and Cue: were launched in a beta form in late 2023, after OpenAI’s ChatGPT popularized generative AI. They served as an experimental laboratory for Naver’s large language model, HyperCLOVA X, with the aim of collecting user data to improve the model’s performance. Now, after more than two years, Naver has decided to end the public services.

Users will not be able to access past conversations after the shutdown, and data recovery will not be possible. Naver told users to back up chats they deem valuable in advance, to preserve any information they wish to keep. The company said the experiments helped confirm the potential of generative AI and that the lessons will help it build a formal AI environment across its search, shopping and other services.

Industry observers say the move reflects a profitability challenge in consumer chatbot services. Building and operating large conversational AIs requires substantial data processing, electricity, and high-end GPUs. Analysts note that while global tech giants have monetized chatbots successfully, regional players like Naver face structural hurdles in sustaining such operations at scale.

Rather than pursue a standalone consumer chatbot, Naver has increased investment in AI data centers and related infrastructure. The company reported spending 1.1595 trillion won last year on servers and hardware, with a sizable portion directed toward Nvidia GPUs. Naver says it is expanding cloud data centers, office space, land and equipment to support its AI infrastructure.

Going forward, Naver plans to embed AI more deeply into existing services such as search and shopping, rather than maintaining separate chatbot offerings. CEO Su-yeon Choi indicated at an earnings call that the company will focus on integrating AI into core services, rather than launching a public-facing chatbot platform.

For U.S. readers, the development matters for several reasons. It highlights how non-U.S. tech firms are pursuing AI by building strong AI infrastructure and integrating AI into commerce, rather than competing primarily through consumer chatbots. It also underscores demand for AI hardware, particularly GPUs, and the financial and energy costs of running large models—issues that shape global AI supply chains and pricing. Finally, it suggests potential cross-border collaboration opportunities and competitive dynamics as international firms scale AI-enabled services that touch global markets.