As countries race to define their roles in the artificial intelligence era, South Korea is emerging as one of Asia’s most determined contenders. Its tech giants, from Naver to Kakao to Samsung, are now embracing “agentic AI”, a next-generation form of AI that acts autonomously to complete goals, not just respond to commands.

Kakao’s everyday AI

Kakao, the company behind Korea’s ubiquitous messaging app KakaoTalk, is taking a pioneering approach to agentic AI. According to the Korea JoongAng Daily, it has developed an AI assistant named Kanana that can manage tasks such as trip planning, gift purchases, and scheduling—all within the chat interface. The integration is powered by a partnership with OpenAI, which brings ChatGPT’s conversational intelligence directly into KakaoTalk.

Kanana uses “Tool Call” capabilities that activate Kakao services like KakaoMap, Melon, or Talk Calendar based on user requests. Seung Kim, head of Kakao’s AI Synergy Task Force, told Korea JoongAng Daily this allows users to access services without switching apps. For example, a family planning a holiday with their dog could ask Kanana to find pet-friendly hotels, make a booking, and coordinate itineraries through ChatGPT. Rollouts are set to begin in October.

To power this ecosystem, Kakao has launched the Model Context Protocol (MCP), a framework that standardises how AI agents interact across apps. Instead of each chatbot learning separate APIs, MCP provides a unified structure. Developers can upload MCP-compatible tools to a marketplace called PlayTools, enabling users to install them like modular add-ons.

Naver’s search reinvention

Naver, South Korea’s dominant search engine, is taking a similar route by embedding agentic AI more deeply into its services, according to Korea JoongAng Daily. The company plans to introduce an “AI Tab” by 2027, turning its search portal into a conversational experience rather than a keyword-based one. According to Kim Sang-bum, head of Naver AI Search, Naver is building both a “core agent” that analyses intent and “vertical agents” that perform real-world actions.

At the centre is HyperCLOVA X, Naver’s large language model built on vast local data to ensure linguistic and cultural precision. Unlike Kakao, Naver has yet to partner with a Big Tech firm, though it remains open to collaboration. The company is also advancing multimodal capabilities—combining voice, image, and location recognition—so its AI can respond contextually to actions such as placing orders or managing calendars.

Naver is expanding into mobility too. Through a partnership with Hyundai Motor, it plans to integrate maps, content, and scheduling into Hyundai’s software-defined vehicles. This will let drivers use a voice-enabled assistant for navigation, weather updates, reservations, and personal schedules, according to company briefings.

T Map’s mobility agent

T Map Mobility, a subsidiary of SK Square, is building a full-fledged mobility agent grounded in years of navigation data. The company aims to link its mapping system with SK Group’s telecom, payment, and electric vehicle infrastructure to create an end-to-end travel platform. Chosun Daily reports that users will eventually be able to plan, book, drive, park, and charge vehicles within a single AI-powered experience.

Park So-ha, head of T Map’s Data and Innovation Division, told the press in September that the company is moving from simple route guidance to “intelligent mobility” that anticipates user needs. The goal is to connect cars, public transport, logistics, and everyday services into one seamless ecosystem.

Enterprise automation

In the corporate world, South Korean companies are adopting agentic AI to boost productivity and cut costs. LG CNS, the IT arm of LG Group, has introduced AgenticWorks in partnership with Canadian firm Cohere, allowing companies to create custom AI systems based on internal workflows, according to the Korea Herald. Alongside it, LG launched AX Sync, a suite of pre-built agents for HR, scheduling, translation, and document management. A coordinating “super agent” manages meetings, approvals, and daily briefings.

These platforms are already used at LG Display, where productivity rose by 10%. LG targets a 30% improvement within three years and claims annual savings of over KRW10bn ($7mn) by replacing foreign software subscriptions with AX Sync.

Samsung SDS is also accelerating its enterprise AI portfolio. Its FabriX platform lets clients design specialised agents, including automatic code converters for legacy systems. Brity Copilot serves as a personal assistant for translation, task execution, and resource management, while Brity Automation applies natural-language prompts to complex operations in finance, HR, and logistics. The Korea Herald notes that Samsung estimates up to 70% of routine office work could eventually be automated.

Still, adoption remains in its early stages. Market tracker Grand View Horizon estimated Korea’s enterprise agentic AI market at $61.4mn in 2024 but expects it to surge to $875.6mn by 2030, Korea Herald reports. Globally, Gartner predicts that by 2026, over 40% of enterprises will use conversational AI agents, up from less than 5% in 2023. McKinsey projects generative and agentic AI could add $2.6 to $4.4 trillion in annual value worldwide, with the biggest gains from automating knowledge work.

Samsung and LG are positioning themselves early, betting that ownership of the underlying infrastructure—hosting, orchestration, and governance—will be as critical as the agents themselves. The key challenge will be proving these systems are secure, reliable, and scalable enough to manage the complexity of real business operations.

The broader vision

As agentic AI matures, it could fundamentally redefine how humans interact with technology, shifting from tool-based use to intelligent collaboration. For South Korean innovators like Kakao, Naver, LG, and Samsung, this isn’t just about keeping up with global competitors. It’s about leading the next chapter of the AI revolution, where technology anticipates human intent rather than merely responding to it.

Yet, this rapid transformation comes with trade-offs. Experts warn that the energy demands of large AI models could strain electricity grids and increase carbon emissions unless powered by renewable sources. There are also growing concerns about workforce displacement as automation scales. The challenge for South Korea will be to balance innovation with sustainability—building an AI-driven future that empowers people without compromising the environment or job security.