%41%49
-
KAIST Develops Vector Database Technology for Reliable Retrieval of Up-to-Date Information
When new information is continually added and old information is removed, an AI system may fail to find the correct material even when that material exists. KAIST researchers have developed a vector database technology that preserves the search paths leading to the needed information as data changes. This allows AI to find up-to-date information more accurately and quickly.
KAIST (President Choongsik Bae) announced on October 5 that a research team led by Professor Min-Soo Kim from the School of Computing has developed CONDA (Connectivity-Aware Dynamic Index). CONDA is a dynamic vector database indexing technology that maintains high search accuracy even as data is continuously added and deleted.
Generative AI models do not inherently have access to information that emerges after training. Retrieval-augmented generation (RAG) has been widely adopted to address this limitation. RAG searches sources such as recent news or internal company documents and uses the results in its answers.
For RAG to work properly, the system needs a search technology that can quickly and accurately find the information relevant to a question among vast amounts of data. Many AI services today convert the meaning of documents or images into vectors, which are numerical representations that computers can process. They then link pieces of information with similar meanings and follow these links to find the material they need.
The challenge is that data in real companies and on the internet changes constantly. New documents, news, and product information keep arriving, while older information is modified or deleted. As these changes repeat, the links built between data points can gradually break, a problem the researchers call "connectivity collapse." The needed information is still in the database, but the AI can no longer reach it, and search accuracy drops. In other words, the information itself has not been lost. The search path that leads to it has been broken.
CONDA considers both distances between data points and whether the search paths needed to reach the information remain connected. When new data is added or existing data is deleted, CONDA preserves the links that are important for search. This prevents specific pieces of information from becoming isolated in the search graph.
In experiments where data was constantly added and deleted, CONDA improved search accuracy (recall) by up to 24.5% compared with state-of-the-art methods. It also increased data update throughput by up to 1.90 times.
The team also ran searches and data updates at the same time for six hours on a large-scale dataset containing 100 million vectors. Throughout the test, CONDA recorded the lowest search latency and the highest search accuracy among the compared methods. This result suggests that the technology can be applied to large-scale AI services.
The technology can be used in knowledge search systems where information such as internal company regulations and work documents changes frequently. It can also be used in a range of AI services where data changes continuously, including news search, product search, and recommendation systems. In particular, it is expected to serve as a foundational technology for RAG, in which generative AI searches external information and uses it in its answers. CONDA allows this up-to-date information to be retrieved reliably over extended periods.
The research results are also being applied to a commercial product. CONDA is being integrated into AkasicDB, the database product of GraphAI, an AI data infrastructure company founded by Professor Kim. It is scheduled for commercial release in the fourth quarter of this year.
"The key value of retrieval-augmented generation (RAG) is that it allows large language models (LLMs) to find and use the latest information they were not trained on at the moment it is needed," said Professor Kim. He added that the study is significant because it presents a data infrastructure that lets AI use up-to-date knowledge accurately over long periods, even in real-world environments where data changes constantly.
Darae Lee, who completed a master's degree at the KAIST School of Computing, is the first author of the study, and Professor Kim is the corresponding author. The findings were presented on September 2 at VLDB 2026 (the International Conference on Very Large Data Bases), a leading international conference in the database field.
Paper title: CONDA: A Connectivity-Aware Dynamic Index for Approximate Nearest Neighbor Search over Evolving Data
DOI: 10.14778/3836663.3836694
This research was supported by the Mid-Career Researcher Program of the National Research Foundation of Korea (NRF) and the SW Star Lab program of the Institute of Information & Communications Technology Planning & Evaluation (IITP), both funded by the Ministry of Science and ICT (MSIT).
2026.10.06 View 543 -
KAIST Develops AI Chip That Recognizes Changes in Motion over Time
For a wearable device to distinguish walking from a brief swing of the arm, it must recognize how a movement changes over time. Korean researchers have developed a technology that stacks semiconductor devices with different response speeds, allowing them to recognize both a change that has just occurred and the changes before it. The technology could be applied to small devices that are worn on or attached to the body to analyze movement and physiological signals.
KAIST (President Choongsik Bae) announced on September 29 that a research team led by Professor Jimin Kwon from the Department of AI Systems, together with researchers from UNIST (President Chong Rae Park) and POSTECH (President Seong Keun Kim), has made a solid thin film from a material that contains ions and used it to stack semiconductor devices with different response speeds in multiple layers.
In these devices, applying a voltage moves the ions within the material and changes the flow of current. When the voltage is removed, the ions require time to return to their original state. As a result, the effect of previous signals remains in the device for a short time. This slow movement has generally been considered a disadvantage, but the research team used it to identify earlier inputs.
The main challenge was controlling how long this effect lasts. Until now, such materials have mostly been liquids or soft gels, which made it difficult to fabricate the devices precisely or to stack them in multiple layers.
To address this challenge, the team made the material into a solid thin film, known as a solid-state ionogel, and adjusted the amount of ions in the film as well as its thickness. This allowed them to create devices that respond to signals either quickly or slowly. When the two types were stacked, one layer responded to the most recent signal, while the other responded to signals that had accumulated over a longer period.
The researchers confirmed that the devices could distinguish all 16 patterns created when four consecutive input signals are turned on or off in different combinations. Using the responses measured from the devices, they also built a simulation model of a system that classifies videos of moving handwritten digits. The model achieved validation accuracies above 90% for sequences of moving images played at different speeds.
The team fabricated the devices on both 4-inch wafers and flexible substrates. Their electrical characteristics remained stable for 55 months after fabrication.
The results could be used to develop technologies for processing information that changes over time, such as movement and signals from the body, directly on small devices. Further validation is needed to determine how the technology performs, and how much power it saves, in actual smartwatches.
"These devices can be fabricated on large-area substrates using existing thin-film semiconductor processes, and they can also be stacked in multiple layers," said Professor Kwon. He added that the technology could be developed into AI chips that analyze movement and physiological signals for devices that must operate with low power consumption, such as smartwatches.
Research Professor Haksoon Jung from KAIST is the first author of the study, and Professor Yong-Young Noh of POSTECH and Professor Jimin Kwon of KAIST are the co-corresponding authors. The findings were published in Advanced Materials on June 30.
Paper title: Monolithic 3D-Integrated All-Solid Ion-Gated Carbon Nanotube Transistors With Tunable Ionic Conductance for Multi-Timescale Reservoir Computing,
DOI: 10.1002/adma.202523703
This research was supported by the National R&D Program through the National Research Foundation of Korea (NRF), funded by the Korean government (Ministry of Science and ICT, MSIT) (RS-2023-00281195 and RS-2024-00355384), and by the Korea Planning & Evaluation Institute of Industrial Technology (KEIT), funded by the Ministry of Trade, Industry & Energy (MOTIE, Korea) (RS-2024-00417909).
2026.09.29 View 913 -
KAIST’s World-First Full-Course Robot Marathon Completion Published in Nature
A quadruped robot ran 42.195 km, a distance that is not easy even for humans to complete, without replacing its battery. KAIST researchers have developed this world-first achievement into design principles for long-distance operation, published the results in Nature, and are now moving toward commercializing robots that can be used in real industrial environments.
KAIST (President Choongsik Bae) announced on the September 24 that the long-distance running technology of RAIBO2, a quadruped robot developed by a research team led by Professor Jemin Hwangbo from the Department of Mechanical Engineering, and the results of its actual marathon completion were published in the international journal Nature.
This is the first time that robotics research conducted in Korea has been published in the main journal of Nature. Even globally, it is extremely rare for robotics research to be published in Nature, making this achievement an important demonstration of both the technological capability and academic level of Korean robotics research.
RAIBO2 completed the full 42.195 km course of the Sangju Dried Persimmon Marathon held in November 2024 in 4 hours, 19 minutes, and 52 seconds without charging or replacing its battery. It was the world’s first case of a quadruped robot completing the full course of an actual marathon event.
The core significance of this paper does not lie simply in the record of a robot running 42.195 km. In the global robotics industry, competition has recently intensified over the running speeds and obstacle-negotiation capabilities of quadruped robots and humanoids. However, to operate robots for long periods in real industrial environments, it is necessary to secure not only momentary locomotion performance but also energy efficiency, thermal management, battery performance, and actuator stability.
The research team approached long-distance operation not as a performance issue involving individual components such as batteries or motors, but as a “whole-robot system problem” in which mechanical structure, electrical systems, and locomotion control are closely interconnected.
Quadruped robots lose energy in multiple areas – including actuators, electrical systems, and mechanical components – as they repeatedly move their legs. Improving the efficiency of only one component makes it difficult to achieve both fast and stable locomotion performance and long travel distances at the same time.
The research team systematically analyzed energy losses throughout the entire robot and jointly designed the mechanical, electrical, actuation, and control systems so that energy consumption could be reduced while maintaining high locomotion performance. Through this approach, the researchers presented specific design principles for which factors must be considered together to enable quadruped robots to move stably for extended periods.
In particular, the research team went beyond short-distance laboratory running tests or maximum travel distances estimated through calculations and validated the technology on an actual full marathon course.
During the marathon, the team recorded the robot’s internal voltage, current, temperature, and battery status in real time and synchronized the data with position, speed, and altitude information measured by GPS. RAIBO2 was officially registered as a participant, allowing formal split times to be recorded, while the researchers ran alongside the robot and filmed the entire journey, lasting more than four hours from the starting line to the finish.
While the 42.195 km completion record demonstrated RAIBO2’s performance, the data accumulated throughout the process provided scientific evidence showing how that performance was made possible and how it was maintained in a real-world environment. The research team developed a single impressive demonstration into research results based on system design principles and actual measurement data that can be analyzed and revalidated.
The research team’s next challenge is to transfer the performance achieved with a single research robot into a product that can be used repeatedly in industrial environments.
RAIBO2 is a research robot that was designed and built by hand by the researchers themselves. To develop it into a product that can be deployed in real industrial environments, the same performance must be achieved reliably across multiple robots while also reducing production costs. The product must also meet industrial reliability requirements, including water and dust resistance, operation across a wide temperature range, durability against shock and vibration, and electromagnetic interference and compatibility requirements (EMI/EMC).
Raion Robotics, a KAIST faculty startup that originated from Professor Hwang’s laboratory, is responsible for connecting the research results to actual products. During the development of RAIBO2, Raion Robotics’ manufacturing technology also contributed to maintaining the high energy efficiency and locomotion performance achieved in the laboratory for more than four hours during the actual marathon.
The research team and Raion Robotics developed not only the mechanical and electrical systems, software, and artificial intelligence but also the motor drivers that precisely control the robot’s motors, allowing them to optimize the entire system.
Currently, Raion Robotics is using this systems-design capability to develop commercialization and mass-production technologies for quadruped robots that can operate for extended periods in real environments, going beyond research prototypes. The goal is to reproduce the world-class performance demonstrated by a single laboratory robot across multiple products and enable stable operation in a wide variety of environments.
Professor Jemin Hwang said, “This study goes beyond demonstrating how far a quadruped robot can run. Using actual marathon data, we demonstrated how the entire robot must be designed to achieve both high locomotion performance and energy efficiency,” adding, “We will connect the world-class performance achieved in the laboratory to products that anyone can use reliably in industrial environments.”
Dr. Choongin Lee and students Donghoon Youm and Jeongsoo Park of the KAIST Department of Mechanical Engineering participated in the study as co-first authors, and the research results were published in the international journal Nature on September 23 (local time).
Paper title: A quadruped robot designed to complete a marathon on a single battery charge, DOI: 10.1038/s41586-026-11102-5
Related photos and videos:
Comprehensive photo/video collection on Google Drive: https://drive.google.com/drive/folders/1ye82tTr2dCK2Bt7Ih0hSgWqkgqlVXk6X?usp=sharing
Full video of the marathon completion on YouTube: https://www.youtube.com/watch?v=-HMFoa3g9nA
The entire process of RAIBO2 completing the marathon can be viewed in the YouTube video.
2026.09.28 View 1884 -
KAIST Names Former Samsung Electronics Chairman and CEO Oh Hyun Kwon as Chair of Its Board of Trustees
KAIST announced that its Board of Trustees elected Oh Hyun Kwon, former Chairman and CEO of Samsung Electronics and Chairman of the Orange Planet startup foundation, as its new chair at the board’s 298th extraordinary meeting in Seoul on September 21, 2026. Prior to the election, Kichul Cha, a KAIST trustee and CEO of InBody, had served as acting chair since the term of former Chair Myung Ja Kim ended in late June.
Kwon earned a bachelor’s degree in electrical engineering from Seoul National University in 1975, a master’s degree in the same field from KAIST in 1977, and a doctorate in electrical engineering from Stanford University in 1985. He joined Samsung in 1985 as a researcher at its Semiconductor Research Institute in the United States. In 1992, he led the development of the world’s first 64Mb DRAM, helping lay the foundation for Samsung’s strategy of building a decisive technological lead over its competitors.
An engineer who rose through the ranks to become a leading business executive, Kwon served as President and Head of Device Solutions Division, Samsung Electronics from 2008 to 2011 and as Vice Chairman and CEO from 2012 to 2017. From 2017 to 2020, he held the positions of Chairman at both Samsung Electronics and the Samsung Advanced Institute of Technology. He is credited with helping propel South Korea's semiconductor industry to global leadership.
After stepping back from corporate management, Kwon served as President of the Korea Invention Promotion Association from 2021 to 2022 and as the seventh Chair of Seoul National University’s Board of Trustees from 2023 to 2025. In September 2025, he became the first business executive to chair the Committee on Mid- to Long-Term Strategy Committee, now under the Ministry of Planning and Budget (then called the Ministry of Economy and Finance). He has also chaired Orange Planet since 2021, supporting young entrepreneurs.
His honors include the Gold Tower Order of Industrial Service Merit in 2009, the Korea Best Scientist and Engineer Award in 2014, the Inchon Award in Industrial Technology category in 2012, and the Dasan Technology Award in 1997. He is also the author of Super Gap (2018) and Again, Super Gap (2020).
Kwon has maintained close ties with his alma mater, KAIST. In 2018, for instance, he became the first KAIST alumnus to receive an honorary doctorate from the university, earning the distinction in business administration. He also helped establish the Kwon Oh Hyun Assistant Professor program, which provides annual support for promising faculty members pursuing creative and ambitious research in physics, mathematics, chemistry, biological sciences, and engineering.
Kwon joined the KAIST Board of Trustees at its 297th extraordinary meeting in August before being elected chair at the September meeting.
His term as chair runs from September 21, 2026, the date of the board resolution, through August 19, 2029, when his term as a trustee expires. Drawing on his experience in building Samsung’s technological leadership, Kwon is expected to help KAIST advance innovation in research and education and further strengthen its global competitiveness.(END)
2026.09.22 View 2812 -
KAIST Develops Smartphone AI That Uses Past Solutions to Tackle Similar Problems
A new AI technology has been developed that retains and reuses knowledge, much like writing down a solution in a notebook and applying it to similar problems rather than asking an expert for help each time. Researchers at KAIST have developed a way for small AI models running on smartphones to store and reuse knowledge from a large server model. The approach reduced server calls by an average of 55.61% compared with a non-cumulative approach while maintaining high accuracy, suggesting that mobile AI could make faster decisions with less server support as it encounters similar problems.
KAIST (President Choongsik Bae) announced on September 21 that a research team led by Professor Jae-Gil Lee from the School of Computing has developed CURE (Cumulative Knowledge Reuse), a technology that enables a small AI model on a device to work efficiently with a large model on a server.
Smartphones have limited processing power and memory, so they typically use small, lightweight AI models. These models can handle simple tasks quickly but may be less accurate when identifying complex or unfamiliar images.
Sending every input to a powerful server model can improve accuracy, but transferring data and waiting for a response takes time. It also adds to network traffic and the server’s computational workload, increasing delays for users and operating costs for service providers.
Researchers have explored a collaborative approach in which the on-device model assesses each input first and sends only difficult cases to the server. Without a way to retain the server’s knowledge, however, the device uses each answer once and may need the same help when a similar input appears. This is much like a student who fails to write down the solution to a difficult problem and has to ask for help again when faced with another of the same kind. The team set out to turn these one-time answers into knowledge the device could continue to use.
CURE first checks whether the on-device model can handle an input reliably. If it cannot, the system consults knowledge previously obtained from the server and stored on the device. It contacts the server only if that knowledge is also insufficient. The process follows three steps, moving from solving a problem independently to consulting previous lessons and, when necessary, asking an expert.
For example, an on-device model that cannot identify a car model in a photo may turn to the server for help. CURE uses the server’s prediction and the image’s features to update the device’s knowledge store. When the device later receives a photo of a similar car model, it can use that knowledge to identify it locally.
CURE does not keep a collection of the original photos processed by the server. Instead, it stores a summary of their shared features and differences. This is like noting a car’s distinguishing features, such as its body shape or headlight design, rather than memorizing the entire photo. The stored knowledge can therefore help the system recognize not only images it has already seen but also similar images it encounters for the first time.
The team tested CURE using vision-language models, which connect images with text to understand visual information. These models link what they see to language, much as people do.
The researchers used MobileCLIP2 on the device and EVA-CLIP, which has 18 billion parameters, on the server. Parameters are numerical values within a model that encode information learned during training. They help the model distinguish objects and identify relationships between them.
In tests on a range of image classification datasets, CURE achieved accuracy close to that of an approach that sends every input to the large server model. It also made an average of 55.61 percent fewer server calls than a device-server collaboration baseline that does not retain knowledge from previous server responses.
In end-to-end tests that included communication time, CURE ran up to 2.80 times as fast as the non-cumulative device-server collaboration baseline and up to 3.67 times as fast as the approach that sends every input to the server. Making fewer server calls reduced the time spent transferring data and waiting for responses.
CURE requires no retraining of either the on-device model or the server model. It leaves both models unchanged and adds a separate store for knowledge obtained from the server, making it adaptable to a range of AI models and services.
The technology could reduce repeated data transfers and server computation image recognition. This could mean shorter waits for users and lower operating costs for service providers.
Potential applications also include robots and wearables that need to recognize their surroundings with limited computing resources. Robots operating over slow or unreliable networks, for example, could use previously acquired knowledge to make more decisions locally without waiting for a server response.
The benefits in practice will depend on the device’s processing power and storage capacity, network conditions, and the characteristics of the input data. Further testing across different devices and environments is therefore needed.
“CURE allows a small on-device AI model to remember and reuse knowledge it has already obtained, rather than repeatedly asking the server the same question,” said Professor Jae-Gil Lee. “By maintaining high accuracy while reducing communication demands and response times, we expect it to help smartphones, robots, and wearables use powerful AI models more efficiently.”
Dr. Youngjun Lee, a postdoctoral researcher from the KAIST Institute of Information Electronics, was the study’s first author, and Professor Jae-Gil Lee from the School of Computing was the corresponding author. Co-authors were Doyoung Kim from Amazon, Junhyeok Kang from LG AI Research, and Professor Hwanjun Song from the KAIST Department of Industrial and Systems Engineering. The findings were presented on September 10 at the European Conference on Computer Vision (ECCV 2026), a leading international computer vision conference held in Malmö, Sweden, from September 8 to 12.
Paper title “CURE: Cumulative Knowledge Reuse for Efficient Device-Server Hybrid Inference in Vision-Language Models”,
DOI: https://doi.org/10.1007/978-3-032-37627-5_21
Author information — Dr. Youngjun Lee (postdoctoral researcher, KAIST Institute of Information Electronics, first author), Doyoung Kim (KAIST graduate, now at Amazon, co-author), Junhyeok Kang (KAIST graduate, now at LG AI Research, co-author), Professor Hwanjun Song (KAIST Department of Industrial and Systems Engineering, co-author), Professor Jae-Gil Lee (KAIST School of Computing, corresponding author)
This work was supported by Institute of Information & Communications Technology Planning & Evaluation (IITP) grant funded by the Korea government (MSIT) (No. RS-2020-II200862, DB4DL: High-Usability and Performance In-Memory Distributed DBMS for Deep Learning, 50% and No. RS-2022-II220157, Robust, Fair, Extensible Data-Centric Continual Learning, 40%), the InnoCORE pro-gram of the Ministry of Science and ICT (AI Meta-Scientist, N10260110, 10%), and Samsung Electronics Co., Ltd.(IO251216-14634-01).
2026.09.21 View 2223 -
KAIST Develops Microbially Produced Adhesive Materials from Glucose Instead of Petroleum
KAIST researchers have demonstrated the potential of using microorganisms to produce a bio-based alternative to petroleum-derived materials for hot-melt adhesives. These heat-activated glues are widely used in packaging, furniture, electronics, and automobiles. Using Escherichia coli, the team produced a new polymer from glucose and demonstrated adhesive performance and thermal properties that support its potential use as a hot-melt adhesive.
KAIST (President Choongsik Bae) announced on September 17 that a research team led by Distinguished Professor Sang Yup Lee from the Department of Chemical and Biomolecular Engineering developed a microbial cell factory by engineering E. coli to produce the new aromatic polyhydroxyalkanoate (PHA) polymers from glucose for adhesive applications.
The research team applied systems metabolic engineering, an approach that redesigns a microorganism's overall metabolism to produce a target compound, to re-engineer the metabolic pathways of E. coli. Through this approach, the researchers successfully produced two new adhesive materials from glucose, poly(4HB-co-PhLA) and poly(3HB-co-4HB-co-PhLA), both belonging to the PHA family of biodegradable polymers.
Adhesives are essential materials used in a wide range of industries, including packaging materials, electronics, automobiles, and construction materials. In particular, hot-melt adhesives (HMAs) are solid adhesives that are melted with heat for bonding and then solidify as they cool. Because they do not require separate organic solvents and enable rapid bonding, they are widely used in industry.
However, most hot-melt adhesives currently in use are based on petroleum-derived polymers such as ethylene-vinyl acetate (EVA). Although they offer excellent adhesive performance, most do not readily degrade in nature and can cause waste and microplastic problems. In particular, even when biodegradable packaging materials are used, if the adhesive does not degrade, the biodegradability and recyclability of the entire product may be reduced.
To address this problem, the research team focused on PHA, a type of biodegradable plastic produced by microorganisms. PHA can be produced by microorganisms using renewable feedstocks such as glucose. It also has the advantage that properties such as polymer flexibility, strength, and heat resistance can be adjusted in various ways depending on which components are included and in what proportions.
Using this principle, the research team incorporated 4-hydroxybutyrate (4HB) and phenyllactate (PhLA) into a single polymer. 4HB contributes to the material’s softness and adhesion, while PhLA increases the material’s rigidity and heat resistance. When the research team varied the relative proportions of the two components, the thermal properties and adhesive performance of the polymers also changed. In particular, polymers containing approximately 24–34 mol% 4HB showed excellent adhesive performance.
The research team designed intracellular metabolic pathways so that E. coli consumes glucose, produces 4HB and PhLA on its own, and then links them together into polymers.
To address metabolic imbalances that arise when multiple monomers are produced simultaneously, the team adjusted the strength and timing of gene expression, added a CoA transferase to facilitate the reactions, and used a genome-scale metabolic model to identify and optimize metabolic pathways associated with insufficient precursor supply and production bottlenecks.
As a result, through fed-batch fermentation, in which nutrients are continuously supplied during microbial cultivation, the team successfully produced 10.2 g of poly(4HB-co-PhLA) per liter of culture medium. By additionally introducing a biosynthetic pathway for producing 3HB, the team successfully produced up to 52.8 g/L of poly(3HB-co-4HB-co-PhLA).
The research team also tested whether the produced polymers could be used as hot-melt adhesives.
In a lap shear adhesion test using stainless steel as the substrate, poly(4HB-co-PhLA) recorded a lap-shear strength of 4.58 MPa. This is higher than the 4.20 MPa achieved by a commercial EVA adhesive. Poly(3HB-co-4HB-co-PhLA) also showed performance comparable to EVA in a wood adhesion test.
In addition, poly(4HB-co-PhLA) retained a substantial level of adhesive strength even after repeated melting and rebonding. The incorporation of the PhLA component improved heat resistance.
Furthermore, after treatment with lipase, an enzyme that breaks down fats, surface damage to the polymers was observed, while the molecular weight and overall mass of the polymers also decreased. These results indicate that the newly developed aromatic PHA can be degraded by enzymes.
This study is significant not only for the development of an eco-friendly adhesive but also for demonstrating the potential of metabolic engineering to produce functional materials with tunable properties. Going forward, this approach could be extended into a biomanufacturing technology that tunes polymer properties, such as adhesion strength, flexibility, and heat resistance, to meet specific application needs, and produces such materials using microorganisms.
The approach also makes it possible to produce functional polymers that have traditionally been produced through petrochemical processes, from renewable feedstocks such as glucose instead. This is expected to serve as a sustainable production technology not only for adhesives but for a wide range of functional polymer materials going forward.
Distinguished Professor Sang Yup Lee said, “This study demonstrates that by engineering microbial metabolism, it is possible to go beyond simply producing polymers and directly produce functional materials,” adding, “In the future, by utilizing various non-natural monomers and microbial cell factories, this approach could be expanded into biomanufacturing technologies for the sustainable production not only of petroleum-based adhesive alternatives but also of a wide range of functional polymers.”
The study, co-authored by Ph.D. student Minju Kang, Dr. Youngjoon Lee, and Dr. Gi Bae Kim of the KAIST Department of Chemical and Biomolecular Engineering, was published online on July 20 in the international journal Nature Communications.
Paper title: Synthesis of poly(4HB-co-PhLA) and poly(3HB-co-4HB-co-PhLA) as sustainable hot-melt adhesives using engineered Escherichia coli
DOI: 10.1038/s41467-026-75809-9
Author information: Minju Kang (KAIST), Youngjoon Lee (KAIST), Gi Bae Kim (KAIST), Sang Yup Lee (KAIST, corresponding author)
This research was supported by the National Research Foundation of Korea through a grant funded by the Ministry of Science and ICT. (Grant No. 2022M3J5A1056117).
2026.09.17 View 4128 -
KAIST Robot That Follows Verbal Directions and Checks Its Own Work Takes First and Second Place in International Competitions
A robot that can follow a single verbal instruction, navigate an unfamiliar building, and verify that it has reached the correct destination has earned top honors in two international competitions.
KAIST (President Chungsik Bae) announced on September 11 that a team led by Professor Hyun Myung of the Urban Robotics Lab in the School of Electrical Engineering took first place in an international challenge held in conjunction with ECCV 2026, a leading computer vision conference, and second place in a challenge at RSS 2026, a leading robotics conference.
The team’s technology goes beyond conventional navigation. It enables robots to interpret their surroundings and independently verify whether they have made the right decision.
Both competitions focused on embodied AI—the ability of a robot to understand human language, perceive its surroundings through cameras, make decisions, and act on them. Unlike AI systems that operate solely in the digital realm, embodied AI interacts directly with the physical world.
Professor Myung’s research team joined forces with the Field Robotics Research Section of the Electronics and Telecommunications Research Institute (ETRI) to form the “URL-FRRS” team. The team placed first among 112 teams worldwide in the VLNVerse Challenge, a vision-language navigation competition held at the ECCV 2026 EMR Workshop in Malmö, Sweden on September 9.
The competition tested whether a robot could reach the correct destination based solely on a stated goal, such as “Find the blue sofa in the living room,” or a multistep instruction, such as “Turn right at the end of the hallway and enter the second door.”
When the research team analyzed why existing robots were failing, they found that, surprisingly, the more common cause was not an inability to find the route, but the robot mistakenly concluding it had “arrived” at the wrong location. A robot might, for example, pass the actual destination or mistake a similar-looking room for the target.
To address this problem, the team developed CoRe-VLN, short for Coverage-based Recovery for Vision-Language Navigation. The technology prompts the robot to verify its location whenever it is about to stop, essentially asking itself, “Is this really the destination?”
Once the robot determines that it may have reached its destination, it scans its surroundings again. An AI model capable of interpreting both text and images then checks whether the robot is in the correct room, whether the target object matches the specified color and shape, and whether the object is actually nearby. The process is similar to a person checking a sign or looking around after arriving at a meeting point to make sure they are in the right place.
If the robot determines that it has stopped at the wrong location, it plans a new route and resumes its search. Because CoRe-VLN can be added to an existing navigation system without further training, it gives the robot a self-verification and recovery capability that can be readily applied in unfamiliar environments.
The team achieved an average success rate of 90.7% across the two competition tasks. Although it tied with the runner-up in success rate, it secured first place by submitting its results earlier. The third-place team recorded a success rate of 89.7%.
The KAIST team also took second place in the NaviTrace Challenge at the RSS 2026 OWN Workshop, held in Sydney, Australia, in July.
In this competition, a robot was given only a single photograph and a brief command, such as “Drive into the parking lot,” and had to determine the correct direction of travel.
When entering a parking lot, people instinctively distinguish the entrance from the exit by interpreting road markings and observing the direction of other vehicles. These cues, however, are not explicitly stated in the command. To navigate spaces shared with people, robots must be able to infer such implicit rules directly from the scene before them.
The team addressed this challenge by developing PRISM-Nav, a system in which multiple AI agents divide the work instead of relying on a single AI model to make every decision.
One agent locates the destination, another identifies hazards such as stairs and curbs, and a third recognizes socially significant features such as sidewalks and crosswalks. A final agent integrates their findings and determines the direction of travel. The approach resembles a group where one person checks the route, another watches for hazards, and a third interprets traffic rules before they jointly choose the safest path.
The researchers also streamlined communication among the AI agents. Instead of describing locations through lengthy text or numerical coordinates, the agents mark destinations and hazardous areas directly on an image and share the annotated image with one another, much like handing someone a marked-up map rather than explaining the route verbally. PRISM-Nav can also be deployed in new environments without retraining the AI models.
The team scored 53 points in the competition, finishing second behind a joint team from Nanjing University and FiveAges.
Notably, the team used multiple instances of the relatively inexpensive and lightweight Gemini 3 Flash model to outperform Gemini 3.1 Pro Preview, the more powerful model used by the organizers as a benchmark. The result demonstrated that coordinating several lightweight AI models can deliver better performance and cost efficiency than relying on a single, more expensive model to handle every task.
“We are delighted that the potential of our technology has been recognized in a competition with research teams from around the world. This achievement demonstrates that AI can make decisions and act independently in unfamiliar environments without additional training.” said Professor Myung.
He added, “We plan to extend this approach beyond navigation to tasks such as identifying, picking up, and moving objects, ultimately developing robots that can verify and correct their own decisions.”
As the technology advances, robots may move beyond simply executing human commands. Before acting, they could independently ask, “Is this the right place?”, “Is this the right object?”, or “Is this action safe?” Such capabilities could benefit not only delivery and guide robots but also robots operating alongside people in logistics centers, factories, hospitals, and other shared environments.
Professor Myung’s laboratory also took first place in international challenges held at ICRA 2026, a leading robotics conference, and CVPR 2026, a leading computer vision conference, in June. The laboratory is now expanding its expertise in spatial cognition into the field of physical AI.
This research was supported by the Ministry of Science and ICT and the Institute for Information & Communications Technology Planning & Evaluation (IITP). (END)
2026.09.16 View 2936 -
Single Material Enables Efficient Charge Injection into Two Types of Ultrathin Semiconductors, Paving the Way for Smaller, More Energy-Efficient AI Chips
A new approach could help make future AI chips smaller and more energy-efficient. A KAIST-led research team has used a single material to address one of the major obstacles facing atomically thin semiconductors, the difficulty of efficiently injecting charge. The technology could contribute to next-generation AI and low-power semiconductor devices in which multiple ultrathin layers are vertically integrated to increase device density and performance.
KAIST (President Choongsik Bae) announced on September 16 that a research team led by Professor Joonki Suh from the Department of Chemical and Biomolecular Engineering has developed a ‘universal van der Waals tunneling injector’ based on tin diselenide (SnSe2). The single-material injector efficiently supplies charge to two different types of atomically thin semiconductor channels.
The study was conducted in collaboration with researchers from Yonsei University, the Beijing Computational Science Research Center in China, the Korea Institute of Science and Technology (KIST), Hanyang University, the Ulsan National Institute of Science and Technology (UNIST), and Samsung Electronics.
A transistor is a microscopic switch that controls the flow of electric current. Nearly every semiconductor chip, from those used in smartphones and computers to advanced AI processors, contains vast numbers of transistors.
Transistors can be broadly classified as n-type or p-type. In n-type transistors, ‘electrons’ carry the current, while in p-type transistors, current is carried by ‘holes’, which are empty electron states that act as positive charge carriers. Modern semiconductor chips combine these two types in complementary metal-oxide-semiconductor, or CMOS, circuits, which enable fast operation while minimizing power consumption.
Two-dimensional semiconductors, which can be less than a nanometer thick, have recently emerged as promising materials for making chips smaller and more energy-efficient. Their atomically thin structure also makes it possible to stack multiple device layers vertically, potentially allowing more transistors to be integrated within the same footprint.
A persistent challenge, however, has been injecting charge efficiently into such ultrathin semiconductors. Depositing a conventional metal electrode directly onto a two-dimensional semiconductor can damage its delicate atomic structure during fabrication. The resulting interface can also create an energy barrier that obstructs the injection of electrons or holes.
In atomically thin semiconductors, the ‘entry point’ for electrical charge has been a major bottleneck limiting transistor performance. Complicating matters further, n-type and p-type channels generally require different contact conditions and have therefore relied on separately optimized electrodes.
The research team addressed this problem using a single material: tin diselenide (SnSe2), a layered semiconductor composed of tin (Sn) and selenium (Se). Instead of forming strong chemical bonds with the semiconductor channel, SnSe2 can be placed in contact with it through weak interatomic attractions known as van der Waals forces. This approach preserves the atomically thin channel and creates a clean, uniform interface.
Crucially, SnSe2 provides a different but favorable charge-injection pathway depending on the semiconductor with which it is paired.
When combined with p-type tungsten diselenide (WSe2), SnSe2 forms a type-III, or ‘broken-gap’, band alignment that enables efficient band-to-band tunneling. In this process, charge carriers pass through an energy barrier quantum mechanically rather than having to acquire enough energy to travel over it. When combined with n-type molybdenum disulfide (MoS2), the same SnSe2 injector forms a different junction. An applied gate field reshapes and narrows the electron-injection barrier, allowing electrons to tunnel through it.
In other words, whereas n-type and p-type transistors previously required separately optimized charge-injection contacts, SnSe2 serves as a single, shared injector that adapts to both channel polarities through different tunneling mechanisms.
The resulting performance improvements were substantial. In p-type WSe2 transistors, the SnSe2 injector increased the maximum drive current by more than 1,000 times compared with conventional nickel electrodes. Drive current refers to the maximum current that a transistor can deliver when switched on.
In n-type MoS₂ transistors, the SnSe2 injector enabled steep switching and an on/off current ratio exceeding one billion (109). This means that the device conducts current effectively in its on state while strongly suppressing current in its off state.
The team also fabricated a CMOS inverter, a basic building block of digital circuits that combines n-type and p-type transistors, and confirmed that it operated reliably in response to repeated input signals.
The key contribution of this research is that it demonstrated a single material can efficiently supply charge to both types of ultrathin semiconductors, moving beyond the conventional approach of using separate electrodes for n-type and p-type devices.
With further advances in direct growth, large-area fabrication, and device integration, two-dimensional semiconductors could eventually be stacked in multiple layers to form three-dimensional chips. Such architectures could accommodate more transistors within the same footprint and perform more functions while consuming less power, contributing to the development of next-generation AI processors and ultra-low-power electronic devices.
“This study demonstrates that efficient charge injection, one of the most challenging bottlenecks in monolayer two-dimensional semiconductors, can be addressed using a single material platform,” said Professor Joonki Suh. “With further advances in direct growth and large-area processing technologies, this approach could accelerate the practical implementation of low-power two-dimensional CMOS integrated circuits.”
The study was led by KAIST PhD candidate Hanbin Cho as first author, with Professor Kyungmin Ko from Yonsei University and Professor Joonki Suh from KAIST serving as co-corresponding authors. The paper was published online on August 12 in the international journal Advanced Materials.
Paper title: “A Universal van der Waals Tunneling Injector for Monolayer CMOS”
DOI: 10.1002/adma.74597Authors: Hanbin Cho of KAIST, first author; Professor Kyungmin Ko of Yonsei University and Professor Joonki Suh of KAIST, co-corresponding authors
This work was primarily supported by Samsung Electronics Co., Ltd. It also received support from Korean funding agencies, including the National Research Foundation of Korea (NRF), funded by the Ministry of Science and ICT and the Ministry of Education.
2026.09.16 View 2472 -
KAIST Researchers Uncover How Neurons Deliver RNA to Where It Is Needed
Much like a courier reading an address on a package, neurons must deliver RNA to the precise locations where it is needed. A KAIST research team has discovered how a small chemical modification on RNA acts as a “delivery tag,” helping selected RNAs travel to distant regions of neurons.
The findings provide a basis for investigating RNA “delivery errors” in brain disorders—cases in which RNA is produced normally but fails to reach its intended destination—and may ultimately inform the development of new therapeutic strategies.
KAIST (President Choongsik Bae) announced on September 14 that a research team led by Professor Ki-Jun Yoon of the Department of Biological Sciences has identified a mechanism through which N6-methyladenosine, or m6A, facilitates the transport of specific RNAs to distal axons, the long, cable-like projections of neurons.
Unlike most other cells, neurons have long, highly branched extensions. Axons transmit signals to other neurons, while dendrites receive incoming signals. For the brain to develop and function normally, these structures must grow properly and form appropriate connections.
Because neurons can extend over considerable distances, RNA produced in the cell body may be needed far away, including at the growing tip of an axon. Neurons must therefore identify particular RNAs among thousands of candidates and transport them to the correct subcellular locations.
Just as producing a package does not complete its delivery—the package must also reach the correct address—producing RNA is only part of the process. RNA must arrive at the right place at the right time. However, the mechanisms through which neurons select and transport RNAs over long distances have remained incompletely understood.
The research team found that m6A, a small chemical modification added to RNA, plays an important role in this process. m6A is a modification of adenosine, one of the building blocks of RNA, and is known to influence several aspects of RNA metabolism, including stability and translation. The new study shows that m6A can also function as a molecular “delivery tag” that promotes the transport of specific RNAs to distal regions of neurons.
The researchers first examined mice in which Mettl14, a component of the enzyme complex that installs m6A on RNA, was deleted from developing neurons. In the absence of normal m6A modification, the neurons exhibited impaired axon projection and neurite development. These findings confirmed that m6A plays an important role in the early development of neurons.
The team then used a high-resolution technique known as m6A-SAC-seq to map the precise locations of m6A modifications in RNA from the developing mouse brain at single-nucleotide resolution. The resulting map revealed that RNAs associated with axon development and synapse organization were among those marked by m6A.
The researchers also identified how this “delivery tag” is recognized and connected to the cellular transport machinery. They found that YTHDF2, a protein that recognizes m6A-modified RNA, works with the RNA-binding protein FMRP and the motor protein KIF5C to facilitate the transport of RNA along neuronal processes.
In the delivery analogy, m6A serves as the tag attached to the RNA, YTHDF2 reads the tag, and FMRP and KIF5C help connect the RNA to the cellular machinery that carries it toward its destination.
YTHDF2 has previously been studied primarily for its role in promoting the degradation of RNA that is no longer needed. This study reveals an additional function: in developing neurons, YTHDF2 also helps transport specific RNAs to distal locations rather than simply directing them toward degradation.
The findings offer a new perspective on the study of brain disorders. Research has traditionally focused on whether RNA is produced and maintained correctly. The new results suggest that researchers should also consider whether normally produced RNA is delivered to the correct location within a neuron.
In other words, neurological dysfunction may potentially arise not only from defects in RNA itself, but also from problems with the molecular tags or transport machinery that determine where the RNA travels.
Future studies could investigate whether such RNA delivery errors occur in neurodevelopmental or neurodegenerative disorders and whether they contribute to neuronal dysfunction or disease onset. If such links are established, the findings could eventually support the development of therapeutic approaches that regulate RNA transport and restore the delivery of specific RNAs to their intended locations.
The high-resolution m6A map of the developing nervous system generated in this study may also serve as a valuable resource for examining how RNA modification and intracellular transport change during brain aging and neurodegenerative disease.
“Our study shows that a small chemical modification on RNA does more than regulate RNA stability—it also plays an important role in delivering specific RNAs to the locations where they are needed within neurons,” said Professor Yoon. “We hope these findings will provide a foundation for understanding why RNA may fail to reach its intended destination in brain disorders and for exploring new therapeutic possibilities.”
Bonsang Koo, Dr. Ajeet Kumar, and Huiseon Hwang of the KAIST Department of Biological Sciences contributed equally to the study as co-first authors. The paper was published in Nature Communications on July 30.
Paper title: The m⁶A reader YTHDF2 regulates mRNA transport and axon outgrowth in developing neurons
DOI: 10.1038/s41467-026-76161-8
Author information: Co-first authors (Bonsang Koo, Ajeet Kumar, Huiseon Hwang); Contributing authors (Minyeong Park, Sung-Min Lee, Ki-Heon Lee, Ji-Hoon Moon, Jaewon Yang, Chan-Woo Park, Saebom Lee, Hee-Jung Jo, Woo-il Kim, Eunji Cho, Hagyeong Lee, Chang-Won Lee, Jae-Sung Woo, Jong Hyuk Yoon, Yoon Ki Kim); Corresponding author (Ki-Jun Yoon)
This work was supported by National Research Foundation of Korea (NRF) grants (2020M3A9E403967022, RS-2024-00332454, RS-2024-00440778, and RS-2026-25488918 to K.-J.Y.) funded by the Korean Ministry of Science and ICT, and Future Planning (MSIP), the POSCOScience Fellowship from the POSCO TJ Park Foundation (to K.-J.Y.), the Young Investigator Grant from the Suh Kyungbae Foundation (to K.-J.Y.), the Korea Brain Research Institute (KBRI) Basic Research Program through the Korea Brain Research Institute, funded by the Ministry of Science and ICT (MSIT) (24-BR-02-03 to J.H.Y).
2026.09.14 View 3695 -
KAIST and KIOM Develop a Wearable Microneedle Electroceutical for Personalized Pain Management
A "wearable electroceutical" has been developed that could reduce the need to visit a hospital or rely on painkillers every time pain occurs. When attached to the skin, it modulates pain through electrical stimulation and can be remotely controlled via smartphone even across long distances, such as between Korea and the United States.
KAIST (President Choongsik Bae) announced on September 13 that a joint research team led by Professor Jae-Woong Jeong from the School of Electrical Engineering at KAIST and Dr. Sanghun Lee from the KIOM (Korea Institute of Oriental Medicine, President Sung-Kyu Kho) has developed a wearable electroceutical platform that combines a wireless microneedle (fine needles that adhere to the skin) electroceutical with Internet of Things (IoT)-based remote control technology.
The key feature of this research is that a single small skin-attached device integrates stable electrical stimulation, smartphone-based remote control, and automatic stimulation based on the body's physiological state. Once its efficacy and safety are validated with actual patients, the technology may be used for personalized pain management at home or in daily life.
The need for such an approach is particularly relevant for chronic pain management. Painkillers are widely used to treat chronic pain, but long-term medication use raises concerns about side effects and dependency. In particular, opioid analgesics used for conditions such as cancer pain carry an increasing risk of tolerance and misuse with prolonged use, which has spurred research into "electroceuticals", devices that modulate nerves through electrical stimulation as an alternative to drugs.
However, existing electroceuticals have limitations. Implantable devices require surgery, and skin-attached electrodes may fail to deliver electrical current properly depending on skin conditions such as sweat or dead skin cells. A concentrated current at specific sites can also raise skin temperature or pose a risk of burns.
To address these issues, the research team developed a temperature-responsive, conductive, and adhesive microneedle electrode (a fine needle electrode that conducts electricity well while adhering to the skin).
Because the microneedles penetrate the highly resistive stratum corneum, the device can deliver stable electrical stimulation while reducing the influence of sweat and dead skin cells. The team also coated the electrode with a conductive hydrogel (a gel-like material that retains a large amount of water) so that current spreads evenly rather than concentrating at the needle tips.
In addition to improving electrical performance, the electrode was designed with a built-in thermal safety mechanism. When skin temperature rises abnormally, the electrode's adhesion weakens and it detaches from the skin on its own, helping reduce the risk of skin burns that could occur during electrical stimulation.
Beyond the skin interface itself, the research team integrated the device with IoT-based remote management. Using a smartphone and cloud server, a healthcare provider can control the device's operating time and electrical stimulation in real time or on a scheduled basis, even when located far from the patient. The team confirmed that the device could be remotely controlled even across long international distances, such as between Korea and the United States.
Following future clinical validation, this could develop into a home-based or remote pain management approach in which patients use the electroceutical under medical supervision without needing to visit a hospital.
The platform further extends beyond remote control by enabling automatic operation based on the body's physiological state. Using a photoplethysmography (PPG) sensor (a technology that measures pulse and blood flow changes using light), the team detected pain-related stress states and, based on this, implemented a closed-loop (a method that automatically adjusts treatment while continuously checking the body's condition) therapy function that automatically triggers electrical stimulation.
The researchers evaluated the performance of the platform in both animal experiments and a small-scale human study. In animal experiments, current was delivered more effectively than with conventional gel electrodes, and pain-relieving effects were also confirmed. In a small-scale study involving healthy adults, changes in skin sensory pain thresholds (the level of stimulation at which pain begins to be felt) following electrical stimulation were observed to assess the potential for application in humans.
However, direct analgesic effects in this study were confirmed only through animal experiments. The research team noted that further clinical studies are needed to confirm therapeutic efficacy and the safety of long-term use in actual chronic pain patients.
Professor Jae-Woong Jeong from the KAIST School of Electrical Engineering said, "By combining a stable skin interface with IoT-based remote management, we have expanded the potential for wearable electroceuticals in daily life. We hope that, following future clinical validation, this approach can evolve into a personalized digital healthcare platform that enables pain management tailored to each patient's condition."
Dr. Sanghun Lee from the KIOM said, "We hope this research will be integrated with future wearable acupuncture technologies to contribute to the development of a new non-pharmacological pain management approach that stimulates acupoints through electrical stimulation."
The study, co-first-authored by Heesoo Kim, a PhD student at KAIST, and Dr. Se Kyun Bang from the KIOM, was published in the international journal Nature Communications on August 28th.
Paper title: Wireless IoT-Enabled Microneedle Electroceutical for Personalized and Connected Pain Management, DOI: 10.1038/s41467-026-76527-y
Author information: Heesoo Kim (KAIST, co-first author), Se Kyun Bang (KIOM/UST, co-first author), Sanghun Lee (KIOM, co-corresponding author), Jae-Woong Jeong (KAIST, corresponding author), and 10 others
Demonstration video: https://www.dropbox.com/scl/fo/zrglys1t1c49l5u1xo08e/AJsbboH3sQMFag-6aO6sSSs?rlkey=dnux0jj80bdczws22u0dq7ugx&e=1&dl=0
This work was supported by the National Research Foundation of Korea (RS-2022-NR067853, RS-2025-02218624, RS-2024-00335066), and by the Korea Institute of Oriental Medicine (KSN2511012 and KSN2511013).
2026.09.14 View 3086 -
KAIST Develops AI That Can Sense “That’s Not What I Meant” Without Being Told
A future in which AI can recognize a person’s unspoken “that’s not what I meant” response from brain signals and adjust its behavior on its own is coming closer. KAIST researchers have developed a technology that detects cognitive mismatch between humans and AI through brainwaves, enabling AI systems to revise their actions in real time according to human goals. The achievement is expected to accelerate the shift from AI that follows explicit commands to AI that can infer human intent.
KAIST (President Choongsik Bae) announced on September 11 that a research team led by Endowed Chair Professor Sang Wan Lee of the Department of Brain and Cognitive Sciences (Director of the Center for Neuroscience-Inspired Artificial Intelligence), in collaboration with Microsoft Research Asia (MSRA), has developed Neural Value Alignment (NVA), a next-generation brain–computer interface (BCI) technology that uses human brainwaves to align AI behavior with human intentin real time.
For AI to collaborate naturally with people, it must accurately understand what a person actually wants. Until now, however, AI systems have largely inferred human intent from externally observable information such as speech, actions, or gestures.
The challenge is that the same action can reflect different goals, and the same goal can also be achieved through different actions. For example, when a person picks up a cup, it may be unclear whether they intend to drink from it or hand it to someone else. Conversely, if the person's goal is simply to quench their thirst, that same goal could be pursued through several different actions — reaching for the cup, picking up a water bottle, or asking someone else to bring a drink. Because of this two-sided “goal–action ambiguity,” when AI misunderstands human intent, users often have to correct it through additional commands or actions.
To address this problem, the research team focused on the “prediction error” signals that arise unconsciously in the brain when a person encounters an unexpected situation. In simple terms, the team used the brain’s instant “that’s not what I meant” response when AI performs the wrong action or pursues the wrong goal.
The researchers distinguished between two types of brain responses. The first is reward prediction error (RPE), which appears when AI misunderstands the person’s ultimate goal. The second is state prediction error (SPE), which appears when the goal is correct but the process or method of action differs from what the person expected.
The team measured real-time electroencephalography (EEG) signals from people as they observed AI performing tasks. They found that the brain produced different signals depending on whether the AI misunderstood the goal itself or chose the wrong method while pursuing the correct goal. The researchers also identified distinctive brainwave patterns that appeared when both types of errors occurred simultaneously.
By applying deep learning to these brain signals, the team developed a technology that can determine, from EEG alone, how a person is interpreting the AI’s behavior. In other words, even without a person saying “that’s wrong,” the AI can recognize whether the human brain is signaling that “the goal is wrong” or “the method is wrong.”
The team then proposed a Neural Value Alignment-based human–AI synergy algorithm, which feeds these decoded brain signals back to the AI in real time so that it can correct its own behavior.
When the AI detects an SPE signal, it interprets the situation as “the desired goal is correct, but the method is wrong” and adjusts its action strategy. When it detects an RPE signal, it understands that “the goal itself was misunderstood” and searches again for what the person truly intended.
Simulation results showed that the proposed method adapted more quickly than existing approaches even in uncertain situations, such as when a person’s goal suddenly changed or some human neural feedbacks were missing.
The key significance of this research is that it demonstrates the possibility of AI systems correcting themselves by reading a person’s unconscious “that’s not what I meant” brain response, without requiring the user to repeatedly say “do it this way” or “that’s not right.”
As the technology advances, it could be applied to physical AI robots in homes and industrial settings, allowing them to understand user intent more naturally and adjust their actions accordingly. It could also be extended to autonomous vehicles that quickly reflect driver judgment, medical and rehabilitation robots for patients who have difficulty speaking or moving, and educational AI systems that adapt to a student’s cognitive state.
Ultimately, the study presents a new model of human–AI collaboration, moving beyond AI that acts only when explicitly instructed toward AI that can sense human responses and adjust itself accordingly.
Professor Sang Wan Lee, who led the international collaboration, said, “This research is meaningful because it shows that AI can move beyond inferring human intent only from visible behavioral outcomes and instead directly use cognitive signals generated in the brain during collaboration with AI.” He added, “The technology can be expanded to a wide range of fields where human judgment and AI behavior must be closely connected, including physical AI, BCI, autonomous driving, precision personalized education, medical robotics, and human–computer interaction.”
Miran Lee, Director, Microsoft Research Accelerator at Microsoft Research, said, “This achievement is the result of the ongoing international collaboration between KAIST and Microsoft Research Asia. We look forward to continuing this partnership to develop world-class BCI technologies that enable humans and AI to communicate and collaborate more naturally.”
The study’s first author is Xin Xu, a Ph.D. student in KAIST’s Department of Brain and Cognitive Sciences. Researchers from Microsoft Research Asia, including Yansen Wang, Dongqi Han, and Dongsheng Li, also participated in the study. The research was published online in August 2026 in IEEE Transactions on Cybernetics, an international journal in the field of cybernetics.
Paper title: Neural Value Alignment: Human–AI Collaboration Under Goal–Action Ambiguity
DOI: 10.1109/TCYB.2026.3722605
Another KAIST–Microsoft Research Asia collaborative study on helping AI rapidly adapt to continuously changing environments was presented in June at ICML 2026, one of the leading international conferences in artificial intelligence. The study was led by Niklas Koeppe, a Ph.D. student in KAIST’s Program of Brain and Cognitive Engineering, as first author.
Paper title: Mitigating Plasticity Loss through Architectural Design in Continual Learning
Original paper: https://icml.cc/virtual/2026/poster/61534
This research was supported by the Institute of Information & Communications Technology Planning & Evaluation (IITP), funded by the Ministry of Science and ICT.
2026.09.10 View 2987 -
Vitamin A Derivative Preserves T Cells’ Function as KAIST Points to a New Strategy for Treating Intractable Brain Tumors
Immune cells can become exhausted after prolonged exposure to cancer, gradually losing their ability to attack tumor cells. This phenomenon is particularly pronounced in aggressive brain tumors and can severely limit the effectiveness of immunotherapy. A KAIST research team has now discovered that all-trans retinoic acid (ATRA), a vitamin A derivative, may help prevent such exhaustion and enhance the efficacy of immune checkpoint inhibitors.
KAIST (President Choongsik Bae) announced on September 8 that a research team led by Professor Heung Kyu Lee from the Department of Biological Sciences, in collaboration with researchers from Seoul St. Mary's Hospital, and Konyang University College of Medicine, has found that all-trans retinoic acid (ATRA), a form of active vitamin A metabolite, suppresses the "terminal exhaustion" of CD8⁺ T cells that attack cancer cells within brain tumors, and can enhance the effect of anti-PD-1 immunotherapy. Anti-PD-1 is a leading immuno-oncology drug (immune checkpoint inhibitor).
Glioblastoma is a representative form of intractable brain cancer that frequently recurs even after surgery, radiation, and chemotherapy. Immune checkpoint inhibitors have shown only limited effect against the disease. One key reason is that immune cells that infiltrate the tumor become exhausted after prolonged combat and lose their functional capacity.
Cells that reach a state of terminal exhaustion in particular lose their ability to kill cancer cells, much like soldiers who have exhausted themselves in prolonged combat. In this state, achieving sufficient therapeutic effect is difficult even with anti-PD-1 therapy, which releases the "brake" cancer cells impose on immune cells.
To address this problem, the research team tried to figure out how to prevent immune cells from reaching total burnout, rather than how to revive the cells. Subsequently, the research team focused on the signaling of ATRA, an active vitamin A metabolite known to regulate cell differentiation and function.
Using an in-vitro model that mimicked the hypoxic, exhaustion-promoting tumor microenvironment of brain tumors, the team induced exhaustion in CD8⁺ T cells. Those conditioned with ATRA progressed to terminal exhaustion at a markedly lower rate. These cells also produced higher levels of immune molecules essential for attacking cancer — including interleukin-2 (IL-2), interferon gamma (IFN-γ), and tumor necrosis factor alpha (TNF-α) — and retained their ability to kill brain tumor cells.
The team also confirmed that ATRA activates WNT/β-catenin signaling within CD8⁺ T cells and increases TCF-1βBD, a TCF-1 isoform important for maintaining T-cell function. In simple terms, the vitamin A metabolite turns on a "function-preserving switch" inside immune cells, helping them avoid complete exhaustion even during prolonged combat with cancer cells.
The effect was also confirmed in mouse glioma models. CD8⁺ T cells conditioned with ATRA maintained better immune function within the tumor, and oral administration of ATRA alone also increased both the number and function of tumor-infiltrating CD8⁺ T cells. As tumor burden decreased, survival was also extended. Notably, in a model of recurrent brain tumors, anti-PD-1 immunotherapy alone showed only limited effect, but combining it with ATRA substantially improved tumor suppression and long-term survival outcomes.
In other words, if anti-PD-1 therapy releases the "brake" imposed on immune cells, ATRA keeps their "battery" from running completely dead. By combining the two approaches, the researchers propose a new combination-therapy strategy that could help overcome the limitations of existing immunotherapy.
The team also analyzed publicly available human glioblastoma datasets. In single-cell transcriptomic data from patients treated with anti-PD-1 therapy, CD8⁺ T cells from responders showed higher retinoic-acid-responsive and WNT signaling gene signatures than those from non-responders. Separately, in an immunotherapy-naïve glioblastoma cohort, patients with a higher proportion of retinoic-acid-responsive CD8⁺ T cells showed significantly better overall survival. The team noted that this study does not directly demonstrate ATRA's therapeutic efficacy in patients, and that further clinical research will be needed to confirm appropriate administration methods, dosages, and combination effects with immunotherapy drugs before it can be applied to actual brain tumor treatment.
Professor Heung Kyu Lee said, "Glioblastoma is one of the cancers most resistant to immunotherapy, because immune cells within the tumor readily become exhausted." He added that the study is meaningful for presenting a molecular mechanism by which active vitamin A signaling helps prevent the terminal exhaustion of CD8⁺ T cells while preserving their anti-cancer function. He also noted that future work validating more precise delivery methods or combination strategies could establish this as a new approach for improving the responsiveness of immunotherapy against intractable brain tumors.
The study was conducted with Dr. In Kang, a postdoctoral researcher in KAIST's Department of Biological Sciences, as first author, and Professor Heung Kyu Lee as corresponding author. Professor Jae-Byum Chang from KAIST's Department of Materials Science and Engineering, Professor Sung Ki Lee from Konyang University College of Medicine, and Professor Stephen Ahn from Seoul St. Mary's Hospital also participated in the research. The findings were published on August 26 in the international journal Signal Transduction and Targeted Therapy.
Paper title: All-trans retinoic acid suppresses CD8⁺ T-cell terminal exhaustion and potentiates anti-PD-1 therapy in glioblastoma
DOI: 10.1038/s41392-026-02852-9
This work was supported by National Research Foundation of Korea grants (RS-2023-NR077244, RS-2024-00439735, RS-2026-25509011 to H.K.L. and RS-2024-00352668 to S.A.). This study was also supported by the Samsung Science and Technology Foundation (SSTF-BA1902-05 to H.K.L.), Republic of Korea. This research was also supported by a grant from the Korea Health Technology R&D Project through the Korea Health Industry Development Institute (KHIDI), funded by the Ministry of Health & Welfare, Republic of Korea (RS-2023-00302794 to J.B.C.).
2026.09.08 View 3716