Naver and Kakao's adoption of a 'multi-AI' strategy is a deliberate and timely response to the dynamic AI landscape.
This move is driven by a convergence of four key factors. First is the productivity narrative. Studies, like one from Microsoft on GitHub Copilot, have empirically shown that AI coding assistants can drastically speed up development—in that case, by over 55%. By providing both a general-purpose tool like ChatGPT and a coding-specialized one like Claude, companies aim to capture these efficiency gains immediately.
Second, enterprise governance has become much more manageable. Previously, using external AI models posed significant security and compliance challenges. Now, providers like OpenAI and Anthropic offer crucial features like data residency in specific regions (including South Korea), single sign-on (SSO), and audit logs. This lowers the barrier for companies to adopt these tools while complying with local regulations like the Personal Information Protection Act (PIPA).
Third, the strategy serves as a crucial risk hedge. The AI market is volatile. In recent months, Anthropic faced potential U.S. government usage restrictions, and its Claude service experienced frequent outages and an abrupt pricing model change. These events served as a stark reminder of the dangers of vendor lock-in. A multi-vendor approach acts as a buffer against such uncertainties in policy, stability, and cost.
Finally, the product and ecosystem narrative is pushing companies towards orchestration. Leading enterprise platforms like Adobe are already integrating multiple AI models, setting a global standard. For Korean companies, the ability to flexibly combine and manage internal workflows with various external tools is becoming essential. This multi-AI setup is the foundation for building more sophisticated AI agents and automated workflows in the future.
- Frontier Model: Refers to the most advanced, state-of-the-art AI models available at any given time, like OpenAI's GPT series or Anthropic's Claude series.
- Orchestration: In this context, the practice of integrating and managing multiple different AI models and software tools to work together seamlessly within a company's workflow.
- Data Residency: A requirement that specifies the physical or geographic location where data is stored and processed. It's critical for companies that need to comply with local data protection laws.