Little has been disclosed publicly about the Korean government’s KRW 150tn (USD 101.5bn) National Growth Fund since it was announced last September. But domestic tech investors sense that it transcends mere subsidy support and will signal a structural reshaping of private markets activity. Now their speculation is coming into focus.

Last week, Korea Development Bank (KDB) said it would allocate KRW 7.5bn under the scheme via several sub-funds, addressing themes such as regional tech industry development, ultra-long R&D cycles, and public participation in the financial story around innovation.

The common denominator is a national agenda ­– encompassing but not limited to the National Growth Fund – that aims to build out the local artificial intelligence (AI) and semiconductor industries. The capital is roundly expected to be channelled into various private equity and venture capital schemes. The key questions are how, how much, and what happens next.

“The structure may differ somewhat from what we typically think of as a conventional VC fund. Still, some capital is likely to flow into the VC scene, which should have a positive impact on overall market sentiment,” said Jinsuk Lee, a team leader at government-backed fund-of-funds KVIC, referring to KDB’s latest disclosure.

“Deep tech and AI companies, in particular, are likely to benefit, and valuations may rise. As for overheating, that would likely require inflows of over multiple trillions [of won] per year, which seems unlikely at this stage.”

Jaeyeon Lee, corporate development manager at tech-focused Atinum Investment, added that the National Growth Fund would provide “meaningful” support to VC fundraising, while also helping individual portfolio companies through direct investment and lending programmes. Other revamps will cater to an AI buildout at the other end of the investment life cycle.

“In particular, the introduction of sector specific tracks – for example an AI track – within the KOSDAQ technology evaluation framework is a positive development for the VC exit environment,” she said. “As the listing process for tech-driven companies becomes more streamlined, we see increased potential for smoother exits and stronger cash flow realization.”

AI advantages

It’s by no means a uniquely Korean story. The idea of investment in tech innovation graduating from consumer internet to enterprise software to AI is well established globally. Yet it can be argued that Korea represents an AI opportunity in this narrative that is eclipsed only by China and the US – with less geopolitical baggage.

Several investors contacted for this story observed that Korea has a broader mix of globally relevant heavy industrial and manufacturing segments than most advanced economies. This notably includes a semiconductor industry rivalled only by Taiwan.

Each major industry is championed by one or multiple chaebols, which collectively are seen as more inclined to adopt new technologies than their global counterparts. It suggests a foundational opportunity to develop AI, especially as it is applied to heavy industries such as shipbuilding, aerospace, and automotive.

Oh-Hyoung Kwon, CEO of deep tech investor FuturePlay estimates that Korea is the third largest market globally behind the US and China in terms of AI talent. He expects this to support a faster investment-driven shift toward an AI-based economy, creating potential for Korea to foster an international VC firm on par with Sequoia Capital or Andreessen Horowitz.

“The government is pushing hard to cultivate AI companies and talent. It’s a fraction of what the US and China are investing, but other than those two countries I haven’t seen much movement,” Kwon said.

“There’s also Japan and Germany from a manufacturing and OEM [original equipment manufacturer] perspective. But I don’t see a lot of talent or disruptive start-ups coming out of those two countries.”

Much of the optimism is tied to the election last June of President Lee Jae-Myung, who made AI development a signature campaign pledge.

According to official data compiled by Citi Research, government spending on industrial R&D was set to stagnate around KRW 60tn a year under ousted president Yoon Suk Yeol. The Lee administration wants to lift it from KRW 58tn in 2025 to KRW 78tn in 2029.

More than KRW 30tn is planned for deployment by the National Growth Fund in 2026 alone. That will include KRW 6tn for AI and KRW 4.2tn for semiconductors. Infrastructure finance and low-interest rate loans will be part of the mix.

Who benefits?

There is a general expectation that the largest and most experienced managers will absorb most of the capital bound for PE and VC. However, Nina Jung, an investment director at NextG Investment and a former KVIC executive, noted that first-time fund managers are unlikely to be ignored. This will be driven primarily by spinouts from established GPs.

“The money is there, they have the track records, and they’re ready to build their own companies. So, why not? It’s a good time to do it,” she said.

Furthermore, as indicated in KDB’s deployment plan, there will be an effort to target more tech investment in areas outside of Seoul. This could benefit small, purely domestic investors that operate deep in regional ecosystems.

Seoul-based Ascendo Ventures has established offices in Daejeon and Busan since 2023 to address a perceived regional AI groundswell. Aaron Shin, a founding managing partner at the firm, explained that although consumer categories are still of interest in Korean hotspots like beauty and entertainment, deep tech has gone from about 50% to 90% of the portfolio in the past five years.

“I recently looked into AI software that is applied to the shipbuilding sector, and the customer requires software engineers to be onsite in order to look at real business processes,” Shin said.

“Only in that way can they build practical software, not just abstract-level software. Many software engineers are going down to the provinces to interact directly with customers, and they’re creating a presence there.”

Most of the public capital earmarked for private markets is expected to go to VC. For private equity, it is largely about entering AI and semiconductor supply chains, finding middle-market OEM suppliers in component segments that have not already been dominated by foreign competitors. But they will not appreciably tilt their overall investment activity toward tech.

Sungmin Kim, co-founder, managing partner and CEO of middle market-focused Ark & Partners, said there would be AI acquisition opportunities for Korean PE in a few years after the boom settles. He believes much of this deal flow will come through sponsor-to-sponsor sales as VC firms pursue exits once valuations normalise.

Ark made just such an acquisition in the mobile software space last September, taking a 100% stake in local services marketplace app Soomgo from Atinum, BonAngels, ID Ventures, We Ventures, and IMM Investment.

In Kim’s view, PE firms will otherwise continue to target the AI theme in tech-enabled services and underlying industrial suppliers such as power system providers. This is more familiar territory, but due diligence must still be upgraded to the AI age.

“Five or six years ago, if I reviewed those types of companies, I wouldn’t need to look at the global value chain because they’re local and not capable of going global,” he explained. “But the investment thesis relies on taking these companies global. So, we have to look at the global value chain, and that’s going to be more complex.”

Pressure points

The question of sector knowledge is arguably all the more pertinent in venture, especially from the global LP perspective.

Morgan Stanley Private Equity Solutions was initially attracted to Korean VC due to the depth of the market in terms of exits and ecosystem, as well as the capital efficiency of local businesses. Part of the appeal was also that so much of the population is concentrated in Seoul, consumer internet ideas can spread quickly.

It anchored the debut fund from Altos Ventures in 2014, citing a refreshingly US-style approach to the market. The fund went on to produce three unicorns, two of which have been fully exited.

The strategy leaned heavily into Korea’s viral consumer strengths. So, it’s an open question whether investment teams with this kind of experience are sufficiently skilled to exploit a start-up market that is increasingly leaning toward industrial AI and deep tech.

Pamela Fung, a partner at Morgan Stanley, said her firm is exploring innovation and industrial technology strategies in South Korea, seeing opportunities in both private equity and venture capital. However, the requirements in terms of GP expertise, including engineering backgrounds, have been elevated.

“You can’t just take someone who was good at consumer internet and then pivot them to be a deep tech person. But finding people with a deep tech background isn’t sufficient either. You also need people with a business mindset who can think about deal parameters and commercialization,” Fung said.

“Finding good investors in this space is going to be as hard as finding unicorns, and it will take time for the space to evolve.”

Local LPs are equally sensitive to this issue. However, those embedded in the government AI mandate may be more willing to nurture a gradual upskilling process.

“What really matters, in my view, is how quickly a VC can learn and adapt – those with a fast learning curve tend to stand out,” said KVIC’s Lee.

The greater concern on the GP side is valuations, especially in the infrastructure and supply chain segments that enable scaled AI deployment.

The main issue is that public capital will cover only about half of the government tech agenda, requiring private LPs with lower risk appetites to make up the rest. They are expected to press for allocations to the largest, most established managers, stoking competition and valuations in a relatively small market of scaled deep tech companies.

NextG’s Jung estimated that Korea only has “tens of companies” in deep tech categories that are mature enough to appeal to the country’s biggest VC firms. In the earlier stages, investors are generally confident that valuations will remain rational. But Jung, who has worked in the local VC space since 2007, highlights previous early-stage bubbles in blockchain and biotech.

“It happens in some fields, especially those backed by the government,” she said. “On the good side, more companies are nurtured and becoming global. On the bad side, if the bubble bursts, some LPs many not commit more capital to new VC funds. That happened in the early 2000s in Korea. I hope it doesn’t happen again.”

Finding angles

Jinsoo Lee, a managing director and head of global investment at Shinhan Venture Investment, observed that the surge of capital will not create a traditional bubble as much as a “quality concentration” dynamic.

In this view, capital is crowded into a small cohort of companies with genuine global technical moats and validated demand. The result is a cycle of accelerated price discovery that needn’t be characterised as indiscriminate exuberance. Lee sees it effectively as a re-rating of Korean deep tech assets toward global multiples.

Shinhan Venture aims to tap into the zeitgeist by targeting both core infrastructure and industry-specific solutions. Success is seen as contingent on anchoring execution-oriented deep tech with verifiable enterprise outcomes, avoiding subsidy dependent models, and using government supports as leverage rather than a crutch.

“On infrastructure, Korea’s Public Growth Fund is explicitly tied to large-scale AI computing projects, including GPU clusters and a national AI computing centre, with tens of thousands of Nvidia GPUs to be allocated to start-ups, SMEs [small and medium-sized enterprises], and research projects,” Lee said, referring to the specialised chips that make advanced AI possible.

“For VCs, this level of subsidized compute compresses time-to-product for model start-ups and vertical AI. It should be underwritten as a structural advantage in speed and capital efficiency rather than as a one-off subsidy.”

Others, meanwhile, believe the most immediate deployment opportunities will be found in infrastructure bottlenecks directly addressed by government programmes.

This will include rising demand for software that improves GPU utilization and cost efficiency, energy and cooling services for data centres, advanced packaging and testing systems for semiconductors, and data services around governance and security.

“VCs should follow budget execution logic across compute, power, data, and semiconductor supply chains, rather than chase policy headlines or compromise due diligence just to meet eligibility requirements,” said Kiho Park, CEO of LB Investment, adding that higher competition and rising valuations were inevitable.

“The real overheating risk appears when narrative runs ahead of technical validation and customer traction, and when exit markets cannot absorb the valuation step-up. Policy can certainly accelerate growth, but the ultimate winners will still be determined by global competitiveness, not domestic policy alignment.”