'No longer optional': K-biotech races to embed AI across drug development, manufacturing and logistics

South Korea's pharmaceutical and biotech industry is accelerating what it calls "AX" — AI transformation — embedding artificial intelligence across the full span of business operations, from research and development to manufacturing processes, logistics and regulatory documentation.

AI has moved well beyond a simple productivity tool, emerging as a core driver that dramatically cuts the time and cost of drug development while reshaping corporate structures. The shift is rapidly redrawing competitive lines across the entire biopharma ecosystem.

A sample survey of 44 domestic pharmaceutical and biotech companies, AI specialists and research institutions — conducted by the Korea Pharmaceutical and Bio-Pharma Manufacturers Association's AI Drug Research Institute — found that 47.3 percent of respondents, or 26 people, are already using AI in drug development.

When those considering or planning to adopt AI (23 respondents) are included, 89.1 percent of the total expressed commitment to AI adoption. The leading areas of application were candidate identification and optimization (15 respondents each), compound identification (11) and mechanism-of-action analysis (9) — all focused on overcoming the limitations of conventional R&D approaches.

Among the therapeutic modalities respondents most want to develop, small-molecule drugs topped the list at 49.1 percent, followed by biopharmaceuticals and antibodies at 21.8 percent. Next-generation biotech areas — targeted protein degradation (TPD) at 10.9 percent and antibody-drug conjugates (ADC) at 9.1 percent — are also seeing rapid expansion of AI adoption.

Concrete results from AI-driven drug development are beginning to emerge. Hanmi Pharm used its proprietary AI platform HARP-pSAR to precisely design protein sequences and identify an innovative drug candidate, HM17321, capable of simultaneously achieving weight loss and muscle gain. The compound has entered a Phase 1 trial in the United States.

Dong-a Socio Group built an intelligent drug development platform called AIDDP by combining the capabilities of its IT affiliate DAI and Dong-a ST's research expertise, consolidating the R&D workflow from candidate screening through structural simulation into a single system. Mokam Life Science Research Institute was selected as the lead institution for a government-funded project under the Ministry of Science and ICT, and is now building an autonomous mRNA research and development framework in which AI analyzes antigen sequences in three dimensions and proposes optimal sequences in an integrated manner.

AI specialist IDivine and GI Innovation jointly built a dedicated AI platform, GI-Orus, using about 1.19 billion data points through a government voucher program, and are using it to quantitatively analyze the probability of success for their immuno-oncology drug GI-102 in clinical trials. Veterinary pharmaceutical company Biod is developing a generative AI-based platform for predicting avian influenza virus mutations and designing vaccines through a Ministry of SMEs and Startups "super-gap" project.

The AI transformation push is spreading well beyond drug R&D into manufacturing, regulatory affairs and distribution. Jio-Young, a dominant pharmaceutical distributor, marked its 24th anniversary by declaring a company-wide AX initiative to embed AI-driven workflows across every stage of operations — from ordering and receiving to outbound shipping and financial settlement. The plan integrates AI into warehouse receiving, purchase returns, dispatch instructions, inventory management and delivery route optimization, while converting the financial settlement process to a real-time AI system to maximize efficiency.

Celltrion plans to build a physical AI smart factory at its new active pharmaceutical ingredient plants 4 and 5, deploying autonomous mobile robots, intelligent robotic arms, collaborative robots and manufacturing management software. Until the factories are completed, the company will first automate standardized tasks, then extend AI to higher-value judgment work to maximize production efficiency. Over the longer term, it plans to introduce humanoid robots as the technology matures, aiming to automate even complex, unstructured tasks that currently require human workers.

Chabiotech has begun building a generative AI software-as-a-service platform through the Ministry of Science and ICT's AX one-stop voucher program. The platform automates the drafting of CMC (chemistry, manufacturing and controls) regulatory documents — widely regarded as the biggest bottleneck in commercializing cell therapies.

Despite the industry's strong appetite for AI, a severe shortage of AI specialists remains a critical obstacle. The association's survey found that 67.3 percent of respondents — 37 people — said their drug development departments have no dedicated AI professionals at all. The main reasons cited were insufficient funding and resources (26 respondents) and a lack of AI literacy among existing staff (19). Industry players identified training AI specialists (37), building data-sharing platforms (25) and investment support (23) as the most urgent priorities for the government and related agencies.

"With global big pharma using AI to raise clinical approval rates and sharply cut R&D timelines, adopting AI is a matter of corporate survival," an industry official said. "Government-level openness of data infrastructure, support for proof-of-concept projects, and a talent-development ecosystem must all be in place before Korea can secure a leading position in the global AI transformation race."