Korea's campuses teach founding at scale. Companies are the harder part.

Korea does not lack campus entrepreneurship. By the end of 2024, according to figures from the Korea Institute of Startup and Entrepreneurship Development (KISED) reported by KoreaTechDesk, universities were running 16,697 entrepreneurship courses with 504,459 participants. Students had formed 5,645 startup clubs with 42,340 members. The output of all this activity is thinner. Student-founded companies numbered 2,083 in 2024, and faculty-founded companies 409. Both counts were lower than the year before.

The numbers describe a funnel that is wide at the top and narrow at the bottom. Measured crudely, and this is our arithmetic rather than a KISED statistic, one student-founded company existed for every twenty or so club members. The more telling point is the direction: participation is broad, and company formation has slipped.

What the data cannot say

The KISED series counts companies founded. It does not say how many survive, raise a second round or hire. None of the sources reviewed for this piece publishes a campus-to-durability survival rate, and a feature that claimed one would be inventing it. What can be said is narrower. Exposure to startup activity is widespread, formation is falling, and the people who study the system increasingly argue the gap sits between the two.

That argument was made in July by Tanachai Kulsomboonsin, co-founder of Canvas Ventures International, in KoreaTechDesk’s analysis of the figures. “Traditional university incubation is usually too passive,” he said. He also observed that “the strongest founders often appear before they know they are founders”, which is a point about timing: the useful intervention comes before a student has decided to start a company, not after a pitch deck exists.

Kulsomboonsin runs a venture studio, so he has a stake in the diagnosis. The diagnosis is nonetheless consistent with the data. A course or a club creates familiarity. A company needs a validated market, a defensible position on intellectual property, a sane cost structure and a route to capital, and few of those are taught in a lecture.

Where founders come from

Korean startup founding is also concentrated. A 2021 survey by the tracker CEO Score, reported by The Korea Herald, covered 123 entrepreneurs at 115 startups. Of the 90 who disclosed their education, 43 had attended one of four universities: Seoul National University (18), KAIST (10), Yonsei (9) and Korea University (6). The sample is small and now five years old, so it is an indication rather than a census. It does suggest that the question is less how to spread entrepreneurship across 5,645 clubs than how a few research-heavy institutions convert talent into companies.

The research universities build their own machinery

KAIST and POSTECH have chosen to own more of the process. KED Global reported in 2022 that Startup KAIST had produced 148 startups, 113 founded directly and 35 more under the university’s umbrella. It runs a startup village in Pangyo, with dormitories and offices, open to students, dropouts and outsiders. Its best-known case from that account is Classum, a communication platform that Lee Chae-rin began building as a second-year computer science student and incorporated in 2017. It raised a US$6m Series A in April 2021, led by Pearl Abyss Capital.

The scale has grown since. Herald Business reported in July that 568 startups were founded at KAIST between 2021 and December 2025, of which 24 went public, with a combined valuation of about KRW 22tn. KoreaTechDesk adds that startups linked to KAIST’s research secured KRW 3.5tn in domestic and overseas investment over five years. These are the strongest figures any Korean campus can show. They should be read with care, though: they count companies the university chooses to claim, and the listed 24 are a small fraction of 568.

POSTECH has built the most explicit capital arm. According to KoreaTechDesk, POSTECH Holdings has invested KRW 55.5bn in 171 companies and helped attract KRW 648.4bn in follow-on investment. That is roughly KRW 11.7 in outside money for each won it deployed, again our arithmetic. The figure matters because it shows what a campus can do that a demo day cannot: put its own money in early and signal to others that a company has been vetted.

Seoul National University has gone for place over capital. Its Campus Town incubator had supported 62 startups since 2020 by 2022, according to KED Global, and evaluates them every six months while offering free tax and legal advice. One tenant, the ultrasound-imaging company Aniai, picked it over KAIST because of its Seoul location.

What the state is doing

Government programmes sit on top of the universities. Challenge! K-Startup, an annual competition running since 2016, is hosted by ten ministries, including the Ministry of SMEs and Startups, the Ministry of National Defense and the Ministry of Environment, according to the Korea JoongAng Daily. The Ministry of Education runs its own track, the Promising Student Start-up Team 300+. In 2024 the top prize was up to KRW 30m, about US$22,000. That is a seed grant, not a financing round, and it favours teams that have already formed.

The Ministry of SMEs and Startups has moved to the university itself. It plans to expand its startup-centred universities from 11 to 17 and to launch three Deep-Tech specialised ones in 2026, in areas such as artificial intelligence, biotechnology, semiconductors and mobility, KoreaTechDesk reports. Deep tech raises the stakes on the formation problem. The same report lists longer development timelines, tighter intellectual-property strategy, specialised teams, regulatory knowledge and heavy capital needs as the obstacles. A club of undergraduates cannot clear those on enthusiasm.

For companies past the campus stage, the main public route is TIPS, a private-sector-led programme run by the Ministry of SMEs and Startups. Its Deep Tech track selected the AI company Dalpa in July 2023, according to Asia Economy, offering KRW 1.5bn in R&D funds over three years on top of KRW 1.3bn in seed investment from Spring Camp and others. TIPS is a tool for companies that have already found one.

The venture-studio argument

The missing step is the one between a club and a TIPS applicant. Kulsomboonsin’s proposal is a venture studio: a team that sits inside or beside the university, works with researchers and graduate students on specific opportunities, and helps build the company rather than judge it at the end. “The goal is to de-risk early ventures and shorten the path from concept to market,” Kulsomboonsin said, as quoted by KoreaTechDesk. KoreaTechDesk points to MIT’s Proto Ventures Playbook, launched in June 2025 and built on the same logic, as a foreign precedent.

Korea has pieces of this. POSTECH’s holding company is close to one in financial terms, and Startup KAIST’s village is close in physical terms. What is less visible in the published material is an operating team that stays with a company through its first customers. If the 2024 decline in student-founded companies is the start of a trend and not noise, the ministry’s expansion from 11 to 17 startup-centred universities will test whether more institutions can be taught to behave like KAIST or POSTECH, or whether those two are simply the places where the talent and the capital were already concentrated.

The next KISED release will help answer that. If courses and clubs keep growing while company counts keep falling, the argument for moving money from exposure to formation will be hard to resist.