The South Korean government has shifted the direction of its startup support policy toward selecting AI startups and nurturing them intensively over the long term. The core of the plan is to inject more than 10 billion won (approximately $7.5 million) per company in research and development (R&D) and commercialization funding, and to link support across each growth stage — from attracting investment to securing sales channels and expanding overseas. The target is to produce 20 AI unicorns valued at 1 trillion won (approximately $749.2 million) or more and one decacorn valued at 10 trillion won (approximately $7.5 billion) or more by 2030.
South Korea’s Ministry of SMEs and Startups announced the “AI Startup Beyond Unicorn Growth Strategy,” prepared jointly with the National AI Strategy Committee, on the 11th at DEEP DIVE INTO 2026, a deep-tech startup forum held at the Grand Hyatt Seoul in Hannam-dong, Seoul. The event runs for three days, from the 10th through the 12th.
The heart of the strategy is “selection and concentration.” The government, universities, research institutes and the private sector will jointly identify a pool of promising AI startup candidates, then select companies in the top 20% by investment attraction and hiring performance as “AI early unicorns.” Selected companies are given stage-by-stage performance targets, and follow-on support continues based on whether those targets are met. Companies that reach the final stage receive more than 10 billion won in R&D and commercialization funding.
The ministry is also overhauling its support system so that it goes beyond technology development to generate actual revenue. Under the approach, startup technologies are applied to large companies’ products and production sites to verify performance, and verified products are then linked to purchases. The “government first proof-of-concept and purchase” program, in which the government and public institutions become the first buyers of startup products, will also be expanded to the AI sector.
A cooperation framework will also be built to vertically connect the entire AI industry ecosystem. The goal is to link AI semiconductor, cloud, large language model (LLM) and AI service companies to support joint proof-of-concept projects. In particular, the government will pursue a co-design project that brings AI chip design firms (fabless companies), AI solution startups and demand-side companies together into a single consortium to jointly develop and verify products. Tailored commercialization support linked to demand-side companies will also proceed in parallel across five core industries, including manufacturing and biotech.
The plan also includes measures to support fundraising and overseas expansion. It will supply investment capital at each growth stage through the Next-Generation Unicorn Project and the National Growth Fund, and build country-specific “AI connectors” to link overseas demand-side companies with South Korean startups. It will offer overseas institutions opportunities to test South Korean AI technology, and will also support securing AI talent at home and abroad. The ministry plans to set up a control tower for nurturing AI unicorns jointly with the National AI Strategy Committee, and to form a coalition involving startups, large companies and investors. It will also pursue regulatory improvements and a joint regulatory sandbox program to ease high-uncertainty AI regulations.
| Category | Details |
|---|---|
| Vision | Realize a top-three AI power and a startup nation |
| 2030 targets | 1 AI decacorn, 20 AI unicorns |
| Selection criteria | Top 20% by investment attraction, hiring performance, etc. |
| Final-stage support | More than 10 billion won (approximately $7.5 million) per company in R&D and commercialization funding |
| Three major strategies | Full-cycle support, accelerating innovative growth, promoting entry into global markets |
| Detailed tasks | 12 |
Lee So-young, South Korea’s Minister of SMEs and Startups, said: “For AI startups to grow into global unicorns, comprehensive support is needed — not only technology development but also commercialization through proof-of-concept, investment and market access, and overseas expansion.” She added, “We will provide concentrated support so that companies can emerge in South Korea that grow beyond global AI unicorns into decacorns.”
The strategy stems from the gap between South Korea’s standing in the AI industry and its startup ecosystem. South Korea ranks around fourth in the world on a global AI competitiveness index, but its share of AI unicorns stands at 1.5%, near the bottom among major countries. The underlying diagnosis is that while South Korea has secured technological capability, a bottleneck arises at the stage of converting that capability into large-scale corporate value.
Market observers note that the government’s approach — assigning performance targets from the early stage and linking follow-on support — differs from conventional broad-based startup support. There are expectations that funding can be concentrated on a small number of promising companies rather than spread thin, while critics also point out that the transparency of selection criteria and the objectivity of performance evaluations will determine the policy’s success. Another key question is whether the joint proof-of-concept system spanning AI semiconductors, cloud, LLMs and services can lead to actual purchases.
In the global market, countries are competitively rolling out startup nurture measures to secure AI sovereignty. South Korea’s strategy is seen as an attempt to connect startups at the application and service stages to global supply chains, separate from large-company-led AI infrastructure investment. Country-specific AI connectors and the provision of overseas proof-of-concept opportunities are expected to serve as channels for South Korean startups to overcome the limits of the domestic market.
