인공지능을 활용한 신약 개발 산업 현황 및 직무별 요구역량 연구

Current Status of AI-Driven Drug Development Industry and Required Competencies by Job Role

This study examined Korea’s artificial intelligence (AI)-driven drug development ecosystem, quantified Korea’s relative technology level versus the United States (U.S.=100) for 2024 and 2030 using a three-round Delphi survey, and organized a role–competency framework for workforce development. Ecosystem mapping identified seven actor groups and three recurring bottlenecks: data access and standardization, multidisciplinary talent and collaboration, and validation/commercialization capacity. In the Delphi survey, 20 experts provided paired ratings for 2024 and 2030. In Round 3, the mean technology level was 63.15 for 2024 and 71.25 for 2030. The primary Wilcoxon signed-rank test showed a significant increase in the 2030 rating compared with the 2024 rating (p=0.004), and sensitivity analyses were directionally consistent. Based on these findings, six role domains were organized: project planning, analysis design, database engineering/management, AI target discovery, AI drug discovery, and machine learning operations (MLOps). In addition, role-level key performance indicators (KPIs) were proposed as illustrative indicators linking process and outcome. Overall, this study presents an integrated framework linking ecosystem bottlenecks, technology-gap assessment, and workforce capability building in AI-driven drug development.