The benchmark for the responsible usage of AI technology in the healthcare sector will promote clarity and high standards for technological applications

< AI Guide for healthcare sector published by KAIST, NUS, and Sense about Science. >
The KAIST Korea Policy Center for the Fourth Industrial Revolution (KPC4IR) published 'Using AI to Support Healthcare Decisions: A Guide for Society.' This global guide is designed to serve as a benchmark for the responsible usage of AI technologies, and will promote clarity and high standards for technological applications in the healthcare sector. The guide details what should be considered when making clinical decisions to help reduce the chances of the AI giving false or misleading results.
The KPC4IR presented the guide in collaboration with the Lloyd’s Register Foundation Institute for the Public Understanding of Risk at the National University of Singapore (NUS IPUR) and Sense about Science, a non-profit organization in the UK specialized in science communication, during the 2021 SIG-KDD (Special Interest Group on Knowledge Discovery and Data Mining) Conference on August 15.
AI technology is being widely used in the healthcare sector and has already proved its accuracy and efficiency in diagnosing and predicting diseases. Despite its huge impact on our daily lives in every sector of society, AI technology has some drawbacks and comes with risks, especially due to biased algorithms.
“We focused on the ‘reliability’ of AI applications in the healthcare sector to make all data well represented, in good quality. The technology will eventually innovate to better serve the people’s new demand, especially critical demands for safety and precision in healthcare services. This global guide will help both developers and people’s understanding of the appropriate technology applications,” says Director So Young Kim at the KPC4IR.
The guide, for instance, says that to scrutinize quality and reliability, the source of the data must be clearly known; the data must have been collected or selected for the purpose it’s being used for; the limitations and assumptions for that purpose have been clearly stated; the biases have been addressed; and it has been properly tested in the real world. It also reflects the importance of the representativeness of data that will affect the accuracy of the AI applications.
“By being transparent and demonstrating the steps taken to check that the AI is reliable, researchers and developers can help give people confidence about providing their data,” the guide states.
For this guide, the KPC4IR and its collaborators collected data after working with numerous experts from the Graduate School of AI at KAIST, the Science and Technology Policy Institute in Korea, Asan Medical Center in Seoul, Seoul National University Bundang Hospital, and AI solution companies.

< From left: Director So Young Kim at KAIST KPC4IR, Director Chan Ghee Koh at NUS IPUR, and Director Tracey Brown OBE at Sense about Science. >
AI is moving beyond answering questions and solving problems into an era in which it acts through robots and forms relationships with people. Amid these changes, how will the roles and responsibilities of humans change, and what should we value as we live our lives? KAIST, together with participating institutions from Kyoto and MIT, is creating a forum to discuss human life and the future in the AI era. KAIST (President Choongsik Bae) announced on the 13th of November that the KAIST AI Philos
2026-11-13Even with high-performance storage devices, inefficient data management can prevent them from reaching their full speed. KAIST researchers have developed a technology that improves the process of moving and reorganizing data to reuse storage space, increasing the data-processing performance of next-generation flash storage devices. KAIST (President Choongsik Bae) announced on November 11 that a research team led by Professor Youjip Won of the School of Electrical Engineering identified the so
2026-11-11DAEJEON, South Korea — Governments use them to track phones. Criminals drive them through city streets and send texts that look like they came from your bank. For 20 years, researchers have been building tools to catch these fake cell towers. They have tested those tools mostly against fake towers they built themselves. That loop has left gaps. Researchers at KAIST built a platform called Devilray that reproduces 2,592 distinct fake-base-station behaviors. They ran it against seven existi
2026-11-04Can AI fine-tuning, that once took 10 hours, be completed in just 3 hours? KAIST researchers have developed a technology that selects the most useful training data and distributes it evenly across multiple GPUs. By reducing repetitive learning of similar data, the method shortened training time by up to 69.6% while maintaining nearly the same AI performance. The technology is expected to help develop AI tailored to specific tasks, such as corporate customer service or assisting medical profess
2026-11-02KAIST will host an event where participants can proudly share failed projects and imagine in advance the kinds of failures that may occur in the AI era. From setbacks already experienced to risks that have yet to emerge, participants will explore together what can be learned from failure. KAIST (President Choongsik Bae) announced on October 30 that it will hold the “4th KAIST Failure Society” from Wednesday, November 4 to Friday, November 13 under the theme “The Future and F
2026-10-30