Abstract
Global AI venture investment has expanded rapidly as a share of the overall
VC market. The share of AI in total VC investment value rose from 23.5% in
2021 to 75.9% in the first half of 2026, while its share of deal count increased
more moderately from 20.4% to 36.6%. This divergence indicates that recent
market growth has been driven more by larger deal sizes than by broad-based
growth in the number of transactions.
Investment has also become increasingly concentrated. In the first half of 2026,
the United States accounted for 84.1% of global AI VC investment value,
later-stage deals for 72.8% of stage-identified investment value, and the top 10
investee companies for 67.5%. Among deals of USD 25 million or more, AI
model companies accounted for 61.0%, highlighting the growing influence of
repeated mega-rounds by later-stage frontier AI firms.
Korea shows a different pattern. The absolute scale of AI VC investment has
remained broadly stable, while AI has gained a larger share of the domestic VC
market as the overall market contracted. Since 2025, investment has become
more concentrated in later-stage firms, top investees, and larger transactions.
Unlike the global market, where AI model investment dominates, AI
semiconductors and AI models accounted for 33.4% and 26.8%, respectively, in
Korea in the first half of 2026.
These findings suggest that policy should preserve entry opportunities for new
and early-stage firms while strengthening continuity between initial, follow-on,
growth, and long-term capital. Financing pathways should also reflect Korea’s
industrial and technological base and the different development periods, capital
requirements, and commercialization paths across AI fields.