South Korea’s Financial Services Commission (FSC) has rejected the Insurance Training Institute’s planned equity investment in an artificial intelligence joint venture, ruling that a nonprofit educational body cannot take a stake in a for-profit company under the country’s insurance laws.
The decision, reported by Chosun Biz on October 19, 2026, is one of the first cases of an insurance regulator explicitly barring an industry institution from investing in a commercial AI venture.
The FSC issued its response on October 2, 2026, via the regulatory sinmungo, a government petition and reporting platform used to field and respond to public regulatory queries.
“An equity investment in an AI joint venture is also deemed to exceed the limits of the scope of business. The Insurance Training Institute is deemed unable to pursue the project, regardless of whether it is through establishing a subsidiary or a joint venture investment,” the FSC said.
The regulator cited Article 178, Paragraph 3, of the Insurance Business Act, which defines the institute’s permitted activities as training, education, and related work. An equity stake in an AI company, the FSC said, does not fall within that statutory boundary.
The ruling comes months after the same regulator introduced human accountability standards for financial-sector AI decisions in June 2026, signalling a tightening approach to AI governance across the country’s insurance and financial services sectors.
Read next: South Korea makes humans accountable for financial-sector AI decisions
The Insurance Training Institute had developed AI-powered insurance education tools, including an AI sales coach, an AI test-generation system, and a personalised learning experience platform. It planned to commercialise these products through a joint venture supplying them to domestic and overseas markets.
Under the proposed structure, the institute would have contributed 1 billion won of its own funds to a 2.5 billion won venture, securing a 40% equity stake. The institute argued that holding a minority position differed from setting up a subsidiary to directly run a for-profit business.
It also said it had put safeguards in place, including a put option allowing full recovery of the investment within three months after the next president takes office.
Insurance Training Institute president Ha Tae-kyung has pushed back against the decision, claiming the regulator reversed an earlier position.
Ha said that during a meeting in the first half of 2026, the FSC’s responsible department head told him that minority equity investments in AI joint ventures were matters the institute could decide on its own without direct FSC involvement, aside from amending the articles of association.
The FSC has not publicly commented on the alleged earlier guidance.
Read more: South Korea targets AI insurance fraud as detection systems come under strain
The ruling arrives as insurance leaders across the region commit significant capital to AI. Accenture’s Pulse of Change survey of 218 senior insurance executives, published in January 2026, found that 90% plan to increase AI investment this year. Yet a quarter of those same executives identified skilled talent shortages as the primary barrier to extracting value from AI.
For brokers and intermediaries, the question is who gets to build the AI tools the industry will depend on. If regulatory frameworks restrict how insurance institutions fund AI development, the pipeline of AI-powered training and distribution tools could tilt further towards private-sector insurtechs.
That shift is already visible in funding patterns. Gallagher Re reported that AI-focused companies captured 95% of the US$1.63 billion in global insurtech funding in Q1 2026.
Business demand for AI-related insurance coverage is also outpacing supply. A Geneva Association survey of 600 corporate insurance decision-makers, published in October 2025, found that more than 90% said they needed insurance to protect against AI-related risks, and over two-thirds said they would pay at least 10% more in premiums for that coverage.
New York University finance professor Haran Segram wrote in a September 2026 analysis published by the Center for Strategic and International Studies that AI liability in the insurance sector “is real, it is growing, and it sits on nobody’s books.”
South Korea’s decision could also signal a direction for other Asia-Pacific regulators. Singapore’s Monetary Authority (MAS) has already issued AI risk management guidelines covering insurers and brokers, as explored in our recent report on Singapore’s AI governance rules and what they mean for the industry before 2027.
The FSC’s decision does not prevent private insurers or insurtechs from developing similar products. But it draws a firm line around what insurance-affiliated nonprofit institutions can do with AI, a distinction brokers should watch as these tools become central to how the industry trains, sells, and distributes.
