New Milestone in AI-Based New Drug Development Regulations

The U.S. Food and Drug Administration (FDA) has announced a draft guidance to promote the safe and effective implementation of artificial intelligence (AI) and machine learning (ML) technologies throughout the new drug development process. This guidance, jointly developed by the Center for Drug Evaluation and Research (CDER) and the Center for Biologics Evaluation and Research (CBER), demonstrates the agency's strong commitment to addressing regulatory uncertainties in response to rapidly evolving AI-based technologies. This is more than just an administrative procedure; it marks a significant shift in the entire biopharmaceutical industry, as the existing regulatory framework begins to formally embrace digital innovation. With the global market for AI-based new drug development projected to reach up to $8.6 billion (approximately 11 trillion KRW) by 2026, this guidance will accelerate the establishment of standard specifications.

Introduction of a 7-Step Risk Framework for Reliability Assessment

The core of this guidance is the introduction of a risk-based '7-Step Credibility Assessment Framework' to systematically verify the reliability of AI and ML models. Sponsors must clearly define the specific context of use (COU) for AI applications and demonstrate, through documented evidence, the quality of data and the predictive reliability of the model, based on the model's risk level. This reflects the FDA's intention to strictly manage the criteria for regulatory approval, as the results derived from AI models are directly related to patient safety and drug efficacy. Specific guidelines are provided for submitting clinical data to prevent bias or opacity in algorithms, which is expected to significantly improve the predictability for sponsors.

Regulatory Scope Extends Beyond Simple Discovery

Notably, this guidance focuses on clinical trial design, data analysis, pharmacovigilance, and manufacturing processes, where AI has a direct impact on decision-making, rather than on early target identification or simple new drug candidate discovery stages. While AI used for simple drug screening is excluded due to its low impact on patient risk and regulatory decisions, high-risk algorithms used for patient selection or safety monitoring will be subject to rigorous verification. This regulatory boundary setting can be interpreted as a policy measure to curb the indiscriminate application of AI technology and to encourage its practical application in later-stage clinical trials with commercial potential. As a result, leading companies such as Recursion Pharmaceuticals (RXRX) with commercialized pipelines will face the challenge of revising their R&D strategies to ensure regulatory compliance.

Long-Term Impact on the Investment Market and the Pharmaceutical Industry

With the release of this FDA guidance, global venture capital (VC) firms and institutional investors will be able to refine their valuation models for AI-based biotechnology companies. In the past, when clear regulatory guidelines were lacking, AI-based new drug development technologies were often seen as speculative. Now, whether or not they pass the FDA's 7-step verification criteria will become a key investment indicator. Although there may be increased initial development costs and data quality control costs in the short term to meet regulatory requirements, in the medium to long term, it is expected that the efficiency of the approval process will be maximized, shortening the R&D period. This is likely to serve as a strong catalyst for traditional large pharmaceutical companies (Big Pharma) to form large-scale partnerships with AI-specialized venture companies and to pursue joint research and development.