The artificial intelligence (AI) alliance between SK Group and Nvidia is showing signs of expanding beyond semiconductors into pharmaceuticals and biotech.
Lee Dong-hoon, CEO of SK Biopharmaceuticals, directly said on the 22nd (local time) at BioUSA about the possibility of a partnership with Nvidia, "It is true we have various considerations related to using graphics processing units (GPUs), and we are breaking down the boundaries of thinking and leaving room for diverse innovation possibilities."
On this day, SK Biopharmaceuticals also presented a blueprint to transform AI into a core "operating system" that leads the entire new drug research and development (R&D) process.
◇ SK Biopharmaceuticals' AI strategy begins to resemble the Nvidia-Lilly model
The core of the "borderless innovation" hinted at by Lee is to elevate AI from a tool that assists specific tasks to an operating system that controls the research and development ecosystem itself.
In the short term, the company aims to create an environment where researchers collaborate with AI agents, and in the long term, to establish a system in which multiple AI agents autonomously support the entire new drug development process, including candidate design, research planning, analysis, and operational optimization.
The goal is to boost research productivity by deploying AI across the entire new drug development process, from candidate design and analysis to research planning, development strategy, and operational optimization.
The industry is paying attention to the fact that this strategy is aligned with the direction Nvidia recently set for its pharmaceuticals and biotech business.
In January, Nvidia and Lilly said they would invest up to $1 billion (about 1.5 trillion won) over the next five years to establish a joint AI innovation lab, setting a goal to build a "continuous learning system" in which lab data and AI hypothesis testing run in tandem around the clock. This aligns with SK Biopharmaceuticals' plan to automate repetitive research processes and maximize productivity.
SK Biopharmaceuticals accumulated more than 2,000 compound synthesis data points and central nervous system (CNS) data amounting to 2.3 million pages of FDA application materials during the development of the epilepsy treatment "Cenobamate (U.S. product name Xcopri)." Nvidia, in its collaboration with Lilly, regarded vast research data and new drug development know-how as core assets.
The strong partnership between SK Group and Nvidia adds further weight to this view.
At a recent meeting in Korea between Chey Tae-won, chairman of SK Group, and Jensen Huang, Nvidia CEO, Chey's eldest daughter, Chey Yoon-chung, head of business development at SK Biopharmaceuticals, and Huang's daughter, Madison Huang, senior director at Nvidia, also attended, drawing attention.
Lee Dong-hoon was also effectively the only Korean CEO invited earlier this year to an Nvidia reception hosted by Huang for global pharmaceutical companies, where they interacted.
◇ First step through a tie-up with Insilico… "Asia's technology, U.S. commercialization" in full swing
The first partner in the "innovation that breaks the mold" mentioned by Lee is Insilico Medicine, a Chinese Generative AI new drug development company.
SK Biopharmaceuticals signed a joint research contract worth up to $2.57 billion (about 4 trillion won) with Insilico to discover innovative new drug candidates in the neuroimmunology area of the central nervous system (CNS). This is the first AI drug discovery (AIDD) case realized through the Open Innovation Center (OIC) launched by SK Biopharmaceuticals.
The contract structure is also differentiated from the past. SK Biopharmaceuticals will lead the entire process from target selection to development, and Insilico's advanced platform "Pharma.AI" will be deployed for early discovery. With this, the company plans to cut the candidate derivation period by nearly 50% compared with before.
In particular, key assets such as molecular design data generated during the collaboration, AI prediction verification data, and structure-activity relationship (SAR) data for compounds will all be accumulated by SK Biopharmaceuticals and used to internalize its own AI capabilities.
Lee said, "Insilico is an excellent partner that achieved the world's first human proof-of-concept (Human PoC) for an AI-based new drug," adding, "Combined with the AI infrastructure capabilities held by SK Group, it is a structure that can fully 'win-win.'"
Insilico is also an example of the "East-West Bridge" model that SK Biopharmaceuticals is pursuing. Based on its experience as the only domestic company to develop a new drug in-house and succeed in direct commercialization in the U.S. (Xcopri), the company's vision is to serve as a bridge that identifies excellent technology and pipelines in Asia (East) and makes them successful in the United States (West).
Lee said, "It took us nearly 30 years to develop and commercialize Xcopri ourselves, but now we have a solid infrastructure with U.S. operations running at 100%," emphasizing, "If we insist on only internal R&D, it will take another 10 years, but by using open innovation and AI as twin pillars, we will dramatically increase speed and the probability of success."
As part of this strategy, SK Biopharmaceuticals recently opened a global innovation space, "LinX," at its local subsidiary in New Jersey, the United States (SK Life Science). In cooperation with the Korea Trade-Investment Promotion Agency (KOTRA), the Korea Health Industry Development Institute, and others, it plans to use the hub as a forward base for Korean and Asian biotech companies to enter the U.S. market and exchange information.
Lee added, "This contract, in which Insilico's AI platform technology and SK Biopharmaceuticals' U.S. clinical and commercialization infrastructure create synergy, is a powerful proof of the 'East-West Bridge,'" saying, "We will evolve the 'extended R&D lab' model, which organically uses external technological strength as if it were our own R&D infrastructure, into a growth platform that can be repeatedly applied whenever new targets are identified in the future."