According to the pharmaceutical and bio-industry on the 11th, Hanmi Pharmaceutical was recently designated as a joint research institution participating in a new project for the '2025 K-AI Drug Discovery Preclinical and Clinical Model Development Program,' organized by the Ministry of Health and Welfare. As AI emerges as a core axis of drug discovery strategy, collaborating with medical institutions and academia to secure government projects has become a new survival strategy for pharmaceutical companies. This project aims to establish preclinical and clinical multimodal datasets, with the core focus on 'developing AI software for reverse translational research design' that can integrate and analyze preclinical and clinical data. Hanmi Pharmaceutical plans to rapidly adopt AI software, applying and verifying hypotheses or novel candidate substances proposed by AI in actual drug research processes, and reflecting the results back into AI learning to establish a virtuous cycle research system. "19.2 billion KRW, 12.75 billion KRW, and even NVIDIA GPUs!" Pharmaceutical companies' AI investment rush is exploding. Not only Hanmi Pharmaceutical, but also major pharmaceutical companies such as Kyongbo Pharmaceutical (a Chong Kun Dang subsidiary), SK Biopharmaceuticals, and Samjin Pharmaceutical have all been selected for government AI projects, securing large-scale investments. Kyongbo Pharmaceutical was selected as a performing institution for the Ministry of Trade, Industry and Energy's 'AI-based Target-Specific Linker-Drug Conjugate Manufacturing Autonomous Lab Technology Development' project, receiving 2.4 billion KRW out of a total 19.2 billion KRW project. In collaboration with the Korea Institute of Machinery and Materials and Korea University, they will establish an autonomous laboratory integrating AI and robotics, along with an automated pharmaceutical manufacturing system, by 2029. SK Biopharmaceuticals was selected for the medical and healthcare consortium of the Ministry of Science and ICT's 'AI-Specialized Foundation Model Project,' acting as a key participating company in the Lunit-led consortium. With support for 256 of NVIDIA's latest GPUs, they will be responsible for AI-driven drug discovery and the development of digital twin models (virtual patient-based clinical trial simulations). Samjin Pharmaceutical was selected for the Ministry of Health and Welfare's '2nd Korean ARPA-H Project,' receiving up to 12.75 billion KRW over 4 years and 6 months to build 'Q-DrugX,' a next-generation drug discovery platform integrating quantum computing and AI. "Cannot do it alone!" The government disperses initial infrastructure costs and lowers R&D entry barriers. An industry official stated, "AI in drug discovery is an area that is difficult for a pharmaceutical company to tackle alone. Government-led projects are highly significant in that the government disperses the enormous initial infrastructure construction costs and risks, which individual companies would find difficult to bear, thereby lowering R&D entry barriers." Indeed, establishing an AI drug discovery platform requires tens of billions of KRW for supercomputing resources, large-scale datasets, and securing specialized personnel. While the financial burden is substantial for individual companies to pursue independently, establishing a multi-institutional collaborative system through government projects allows for cost dispersion while accessing top-tier AI technology. The virtuous cycle research system being built by Hanmi Pharmaceutical involves verifying AI-predicted candidate substances through actual experiments and then feeding those results back into the AI for improved accuracy. This is expected to increase the drug discovery success rate from the current 15% to 45%. The drug discovery period can also be more than halved, from 10-15 years to 5-7 years. The sense of crisis that falling behind in the AI drug discovery competition could lead to being marginalized in the global market is driving pharmaceutical companies into the race to secure government projects.