1. AI penetrates the clinical setting: limitations of traditional approaches

The clinical trial to date has been a “static” model that adheres rigidly to a predefined protocol. However, this paradigm does not suit artificial intelligence (AI) that learns and evolves autonomously. Although patient data accumulate in real time, the inability to update algorithms has prevented the immediate translation of cutting‑edge technology into practice—a longstanding challenge.

2. Embedding real‑time data and continuous updates into trials

The research team proposed a new protocol based on a continuous‑monitoring system to ensure that the AI model never stops learning. Therapeutic efficacy and adverse events are detected in real time, and the model is periodically updated once safety is confirmed. Throughout this process, data quality is strictly controlled and an automated risk‑detection system is activated, achieving both “speed” and “safety.”

3. Sustainable AI‑driven medical innovation: a shift in regulatory paradigm

Regulatory agencies now need to establish a “dynamic clinical‑trial framework” that permits flexible AI updates. This will not only facilitate technological advancement but also dramatically accelerate drug development and provide truly personalized medicine to each patient in real time.

4. Implications and outlook

In the future, AI‑based medical devices and therapies will continue to improve after approval, creating an era of “living treatments” that adapt to individual patient conditions. This model, in which technology and regulation evolve together, is expected to become the blueprint for a safer, smarter healthcare system.