OpenEvidence, known as "ChatGPT for doctors," raised $200 million on the 20th (local time), valuing the company at $6 billion. Its valuation, which was $3.5 billion just three months ago, has nearly doubled. This U.S. Start - Up, founded less than three years ago, uses a model trained on leading medical journals such as JAMA and NEJM to help doctors and nurses search and verify medical evidence in real time. Despite a closed structure accessible only to medical staff, monthly clinical queries have surpassed 15 million. Global investors including Google Ventures, Sequoia Capital, Blackstone, and Kleiner Perkins participated in the round.
As competition in general-purpose generative artificial intelligence (AI) reaches saturation, Start - Ups are placing bets on "industry-specific AI (vertical AI)." In a market dominated by big tech that controls large AI models and infrastructure, they are seeking survival strategies with specialized AI focused on specific industries and domains such as healthcare, law, and academia.
According to market research firm Global Market Insights on the 22nd, the industry-specific AI market was about $10 billion (about 14.22 trillion won) last year and is expected to grow to $115.4 billion (about 164.2 trillion won) by 2034. It is still only a portion of the overall AI market, but its compound annual growth rate (CAGR) reaches 21.6%.
According to a survey by Boston Consulting Group (BCG) in Aug., among 1,250 corporations worldwide, only about 5% have generated tangible business value through AI investment. However, many of these are corporations that adopted AI optimized by industry, such as healthcare, law, and manufacturing, and BCG said a tailored approach is needed for AI to take hold as a practical "work tool."
In the legal sector, Harvey and Hebbia are leading the market. Harvey is an AI assistant that supports contract review, document automation, and regulatory response for global law firms, and in June it raised $300 million in a Series E round, reaching a $5 billion valuation. Since 2024, it has been deployed at the U.K. law firm Ashurst and Verizon Communications Inc. in the United States, and is being used in actual legal review systems. Hebbia is an analytics platform that allows natural language queries across large documents, spreadsheets, and reports, and in a 2024 Series B round it secured $130 million, valuing the company at $700 million. Recently, it partnered with financial data analytics company FactSet to launch a solution that integrates analysis of unstructured documents and structured data, expanding into the financial market.
Such industry-specific AI is spreading rapidly beyond healthcare and law to other sectors such as science and manufacturing. Lila Sciences in the United States, which champions AI for scientific research and drug development, recently raised an additional $115 million from investors including Nvidia. Its valuation stands at $1.3 billion (about 1.8486 trillion won). It is regarded as a platform that integrates experimental data and papers to serve as an "AI experiment partner."
In Korea, NCSOFT is cited as a central axis of the spread of specialized AI. NCSOFT is a leading Korean game company and is expanding its AI business beyond games through NC AI, a subsidiary dedicated to AI research. NC AI was selected for the government's "national AI team (independent AI foundation model project)" alongside LG, SK Telecom, and Naver, standing shoulder to shoulder with large corporations. It is also participating as the lead organization in the "K-content AI innovation leadership project" promoted by the Korea Creative Content Agency (KOCCA), expanding AI application across the content industry, including media, fashion, and music. NCSOFT's multimodal model "VARCO" series features an architecture that processes language, vision, voice, and 3D data in an integrated way, representing an example of extending AI technology accumulated in game development to industrial use. Among them, "VARCO VISION 2.0" is a vision-language model (VLM) that can understand images and text simultaneously to analyze documents containing tables and charts, and it can be used across various industries, including finance, education, culture, and manufacturing.
AI search agent Start - Up Liner is drawing attention as an AI for academia and research. Liner is the only Korean corporations to be included for two consecutive years in the "Global AI Web Service Top 100" selected by global venture capital Andreessen Horowitz (a16z), and its cumulative users grew more than 10% in half a year, from 11 million in Feb. to 12 million in Aug. Recently, it launched a "research AI agent" for researchers, offering academia-specific Generative AI features such as idea generation and citation source recommendations. Liner is also participating in the national AI team consortium led by SK Telecom, contributing to improving AI accuracy by providing real-use datasets and sentence-level confidence verification technology.
An AI industry official said, "Now not only Start - Ups but also large domestic corporations recognize the reality that it is difficult to beat big tech in the general-purpose AI race," adding, "the shift of gravity toward specialized AI that can leverage strengths in each domain and dataset is inevitable."