"We are developing AI to treat everyone," said Hwang Tae-hyun, Endowed Director of the Artificial Intelligence Research Institute at Vanderbilt University School of Medicine in the United States, on Wednesday. "When we observe tumors through AI, conquering all cancers becomes possible."
Hwang made the remarks at Session 2, "Redesigning Life: Redrawing the Blueprint of Life with AI," held at the Shilla Hotel in central Seoul as part of Seoul Forum 2026.
Hwang has built a system in which four agentic AIs work together organically for cancer treatment. The system observes drug interactions at the cellular level inside the body, captures the mechanisms by which cancer cells evade drugs, and identifies the optimal drug combinations. The AI's role is to help doctors make the most appropriate decisions for each individual. The system has actually been used to discover a new targeted anticancer drug for a patient diagnosed with stage 4 breast cancer.
Through the U.S. cancer-conquering project "Cancer Moonshot," Hwang has joined forces with Harvard University, Columbia University, and Korean institutions including Asan Medical Center, Severance Hospital, and St. Mary's Hospital. He also collaborated with the LG AI Research Institute to create a pathology AI that predicts cancer gene activity within tissue in less than one minute. "We developed it not for research purposes but to be used clinically within a year," he said.
Han Nam-shik, head of the AI Research Center at the Milner Therapeutics Institute at the University of Cambridge and a professor in the Department of Quantum Information at Yonsei University, emphasized that AI can reduce the astronomical costs and time required for new drug development. "Developing a new drug requires on average about four times the cost of developing a rocket, but the success rate is very low, around 10%," Han said. "AI is the breakthrough that will overcome these limitations."
Han analyzed that combining quantum technology with AI in new drug development would maximize synergy, in addition to AI alone. The approach applies the principle of superposition, a core concept of quantum mechanics, to AI algorithms to enhance search efficiency. In a verification using SARS-CoV-2 data with an actual quantum walk, Han's research team successfully identified key protein modules deep inside cells that conventional methods had failed to find. "It has been proven that using AI and quantum technology enables analysis of large-scale datasets in two and three dimensions, allowing for deep exploration," Han said.
At the forum, suggestions were also made to advance the pharmaceutical and biotech industries powered by AI. Voices called for active regulatory innovation. "Korea has good patient data, but hospitals and institutions hold the data separately and do not share it," Hwang said. "Just stockpiling data has no value. It is meaningful only when the data is processed into a form that AI can actually use, with patient information serving as input values."
The need for cultural improvement was also raised. "Korea has a strong craftsman spirit, but organizations and institutions tend to divide data among themselves and not collaborate," Han said. "It is regrettable that there is a lack of substantive convergence research and a culture of data sharing." Park Tae-yong, Vice President of Galux, said, "I believe that if we move forward as one team with an open innovation mindset, we can build a competitive pharmaceutical and biotech ecosystem."
