JRX EnerTech Solutions, a specialized battery foundry company, is accelerating its growth by targeting high-mix, low-volume battery markets such as energy storage systems (ESS), robotics and drones. Revenue surged over fourfold from 4.9 billion won in 2024 — the first year of plant operation — to 20.1 billion won last year. The company targets 32 billion won this year and aims for over 100 billion won next year.

JRX EnerTech Solutions focused on market segments with granular product specifications and diverse demands — areas difficult for major battery manufacturers to address via large-scale dedicated production lines. When battery fabless companies focus on product development, JRX EnerTech Solutions handles mass production, including electrode manufacturing and cell assembly.

The number of corporate clients seeking foundry services is growing rapidly. Currently, the company maintains clients including three automotive OEMs, South Korea's top three battery makers and approximately 40 battery fabless firms. Notably, five clients that collaborated from the development stage have transitioned into mass production.

Increased mass production volume directly led to capacity expansion. JRX EnerTech Solutions is currently operating its 500 megawatt-hour (MWh) electrode production line at its Eumseong plant in North Chungcheong Province at full capacity. By installing a 1.5 gigawatt-hour (GWh) second unit by the end of this year, total production capacity will expand to 2GWh by February next year. Cell assembly capacity will also expand from the current 100,000 cells to 300,000-400,000 cells.

Oh Duck-geun, CEO of JRX EnerTech Solutions, said, "This year served as the year we validated the foundry business model through client expansion, mass production conversion and technological diversification. As five development clients transitioned to mass production, we finalized the decision to expand production capacity."

The foundry's reach is also expanding into after-sales (AS) service batteries for electric vehicles (EVs). Even after an EV model is discontinued, replacement cells must be supplied during the warranty period. Because batteries cannot be stored long-term as inventory, they must be manufactured on-demand. Instead of restarting legacy production lines, major battery manufacturers can outsource this work to a foundry specialized in high-mix, low-volume production.

JRX EnerTech Solutions is currently negotiating AS battery production with South Korea's top three battery makers. One partner is reviewing approximately 100 projects for AS production conversion. By supporting production during the development stage of next-generation batteries and taking over AS volume after product discontinuation, the company is broadening its touchpoints with major battery makers.

JRX EnerTech Solutions plans to expand its domestic production capacity to at least 6GWh. As discussions progress regarding approximately 4GWh of volume with a U.S. ESS client, plans for additional investments near the Eumseong facility are accelerating. The company is also preparing for an initial public offering (IPO) to secure funds required for capacity expansion.

Oh added, "Based on our planned production capacity, we project achievable revenues of around 200 billion won, and we expect over 100 billion won to be feasible next year. We are preparing to submit our listing application in March 2027, targeting an official IPO in September of the same year."

Q&A

Q: Hello, this is Lee Soo-hwan from THE ELEC. Today, we are joined by CEO Oh Duck-geun of JRX EnerTech Solutions to discuss the latest updates on the battery foundry market and industry conditions. Hello, CEO Oh.

A: Hello, I am Oh Duck-geun, CEO of JRX EnerTech Solutions.

Q: We've been having regular check-ins throughout the year. Over the past few years, while the overall battery market has been expanding, domestic companies haven't captured that growth, leading to various industry issues. People talk about slowing EV sales, but looking at Tesla Model Y sales in Korea, that doesn't seem entirely true. Ultimately, the market is growing, but domestic firms failed to capture it. However, your battery foundry business wasn't targeting that main EV segment from the beginning, right?

A: When we launched the foundry business, our initial target focused on non-EV segments — specifically ESS, defense, drones and robotics, which are receiving tremendous attention right now. Ultimately, we are also looking at the EV market, but rather than mainstream production, our focus is on the AS market. When an EV is sold, the battery carries a 10-year warranty. Auto manufacturers and battery makers face a structural issue where they must continue producing cells throughout that 10-year window. Demand for someone to handle that operational burden is surging. The opportunity we see in the foundry market isn't the core mainstream market dominated by the top three battery makers, but the AS market — and those markets are opening faster than expected. We have already initiated discussions with domestic battery majors and auto OEMs regarding AS supply. The defining characteristic of the AS market is precisely high-mix, low-volume production, which aligns with our core focus.

Q: Exactly. AS volume doesn't ship at a constant pace, nor can inventory sit indefinitely.

A: Because battery capacity and quality degrade over time, you cannot produce and stockpile inventory; it must be manufactured on-demand. Major corporations excel at low-mix, high-volume manufacturing, whereas high-mix, low-volume production must be handled by a dedicated foundry. While we don't treat the main EV market as a primary focus today, we see long-term potential, and the target markets we identified are already opening up. The foundry business is growing much faster than we originally anticipated.

Q: Could you highlight a few of your representative client groups?

A: While I cannot mention specific client names directly, I can describe our primary client segments. First, we are engaged in foundry operations with automotive OEMs. Auto manufacturers are currently developing proprietary cell designs internally, and we provide foundry execution for those R&D efforts.

Q: Do these include domestic automakers?

A: Yes, both domestic and overseas automakers. Orders are steadily increasing, so we view this very positively. Our second client segment is South Korea's top three battery manufacturers. People often ask why major domestic battery makers — who possess massive capacity — would outsource foundry work to JRX EnerTech. We view our relationship with the top three through a two-stage lens: front-end and back-end. The front-end involves supporting their R&D. Previously, when the focus was strictly on high-nickel chemistry, they only developed high-nickel. Now, they must simultaneously develop lithium iron phosphate (LFP), sodium-ion and all-solid-state batteries. It is extremely difficult to process all these next-generation chemistries on existing R&D pilot lines, so they leverage us to accelerate development. That is the front-end. The back-end is the AS market I mentioned earlier. The relationship between major battery makers and a foundry isn't competitive — it's mutually complementary.

Q: For example, if LG Energy Solution has a line sitting idle to reduce fixed costs, restarting that factory just to produce a single month's worth of required cells incurs massive operational and manpower costs. Running it for one month and pausing again results in a net loss. Outsourcing that volume to JRX makes far more financial sense — and that calculation is driving the AS market.

A: Exactly. That market always existed, but it remained inactive because entering the automotive supply chain requires a foundry partner that is thoroughly vetted for reliability, scale, quality standards and financial stability. Historically, such players didn't exist, so the top three didn't look outside. Now that a competitive foundry player like JRX has emerged, they are actively pursuing it.

Q: As electric vehicles become mainstream, AS demand is compounding, and older EV models are accumulating. Outsourcing those runs is far more efficient than internal manufacturing.

A: Correct. Those legacy production lines are less efficient, run sporadically and consume valuable floor space. Battery majors want to decommission legacy lines and install new high-speed equipment. That operational need aligns perfectly with our capabilities, and formal discussions are now underway.

Q: Aside from automotive, what other major markets are you expanding into?

A: Our third core client segment consists of battery fabless companies.

Q: By fabless, do you mean emerging battery startups?

A: Emerging battery companies that focus heavily on core R&D and market development. What they lack is mass production capabilities. They often lack commercialization and mass manufacturing expertise, so our core foundry model centers on strategic partnerships with them. This client base is expanding rapidly.

Q: How many active clients do you currently have?

A: We have three automotive OEM clients, all three domestic battery majors and approximately 40 core battery fabless partners.

Q: That already totals nearly 50 clients. It sounds like your Eumseong plant is already operating at full capacity.

A: We are operating at 100% capacity. Our 500MWh cathode and anode lines are fully set up and running. Client demand is strong, and growth will accelerate next year. That is why we are currently preparing our second line.

Q: Will Line 2 be installed within the existing plant? What does the process layout look like? Are you referring to post-mixing processes?

A: The process layout is identical. We currently operate one complete set covering mixing, coating, pressing and slitting. We are bringing in an identical second set, but this line is specifically optimized for mass production. Line 1 was semi-automated for pilot-scale flexibility, whereas Line 2 is an automated mass-production line arriving at the factory by late this year. It will be fully operational by February next year and contribute directly to 2027 revenue.

Q: Combined, Line 1 and Line 2 will total 2GWh of capacity?

A: Correct. Our current capacity is 500MWh, and the new line adds 1.5GWh, bringing JRX's total production capacity to 2GWh by February next year.

Q: Is the 1.5GWh expansion allocated for EV mass-production clients?

A: It is not for EVs. This year served as the validation year for our foundry business model. First, our client base expanded past 50 companies. Second, clients that previously requested pure R&D are now transitioning into commercial mass production. That transition created the need for Line 2. Third, our technical portfolio diversified significantly. Specifically, five client projects successfully transitioned from R&D into full mass production, which justified adding the 1.5GWh capacity.

Q: So fabless clients that secured end-customers and required large cell volumes are now moving into mass production.

A: Exactly. Our core purpose is helping clients commercialize, mass-produce and sell the products they developed. We now have five clear success stories. Looking at their end-customers, these strategic clients manufacture specialized cells for ESS, logistics robots, drones and special-purpose IT/wearable devices. As they enter mass production, JRX is scaling its capacity to 2GWh to support them.

Q: Are there other client categories?

A: We also partner with battery material developers. Companies developing novel materials find it difficult to convince cell manufacturers using material samples alone. Consequently, many adopt a strategy of building complete cells first to prove performance in the market. Waiting for domestic battery majors or Chinese firms to adopt a new material takes years. To bypass that delay, material developers use JRX to manufacture complete cells containing their materials, selling them into specialized markets for rapid validation. Because they lack manufacturing infrastructure, they collaborate heavily with JRX.

Q: How are your revenue metrics trending?

A: We founded the company in December 2022 and officially brought the factory online in February 2024. In 2024, we prioritized operational stabilization over aggressive sales, recording 4.9 billion won in revenue.

Q: In 2024.

A: In 2025 — our first full year of commercial operation — revenue reached approximately 20.1 billion won.

Q: Passing 20 billion won means a fourfold increase year-over-year.

A: Yes, roughly fourfold growth. This year represents the full validation of the battery foundry model. Response from clients has been exceptionally strong, and as of August, our full-year 2026 revenue is virtually locked in at approximately 32 billion won.

Q: What about profitability?

A: Achieving early profitability in the battery industry is notoriously challenging. We posted net losses in 2024 and 2025, but we are executing a complete turnaround to profitability this year. We expect to post solid positive numbers.

Q: What operating margin are you projecting?

A: Ongoing capital investments generate substantial depreciation expenses, making high margins difficult in the short term. However, we are targeting an operating profit margin of approximately 10% this year, and we expect to hit that target.

Q: Looking back over the past year, JRX signed numerous corporate partnerships, including Scandinavian newcomer Morrow Batteries and Australia's 1414 Degrees. Which partnerships stand out as the most significant?

A: Collaboration is the foundation of our business model, and our partnerships fall into three main categories.

First are client partnerships. We have established formal MoUs with key front-end clients like Amprius, Nyobolt and ElectroActive, and we have moved past initial MoUs into concrete negotiations for mass production volumes.

Second are material suppliers, such as 1414 Degrees and Norwegian synthetic graphite maker Vianode.

Third is a novel model: leveraging idle global production capacity. Morrow Batteries was our first partnership in this category. As European battery firms faced operational hurdles after building gigafactories, idle capacity emerged. We initiated discussions to utilize those facilities on behalf of JRX's client base. At the end of September, a major French manufacturing firm will visit our facilities to sign an MoU focused on JRX utilizing their idle production lines.

Q: So based on your track record of rapidly stabilizing your plant, these global firms are asking JRX to help stabilize and run their gigafactories?

A: It's a slightly different strategic approach. They want JRX to route our client orders into their facility to utilize their idle capacity. From our perspective, while we continue investing domestically, our footprint is currently limited to South Korea. Rather than spending capital to build new plants in Europe, we use these partnerships as our European operational hubs.

Q: It functions almost like a "foundry for foundries."

A: Exactly — a foundry for foundries. Directly committing capital expenditure (CAPEX) overseas requires immense capital, time and complex European permitting processes. Partnering allows us to bypass those hurdles. Our global strategy centers on our Eumseong facility, which will expand to 2GWh, while we actively discuss a roughly 4GWh project with a U.S. ESS client.

Q: Regarding the 4GWh project, plans for a second domestic site near Eumseong are taking shape.

A: Timeline planning is accelerating, and we are evaluating site options simultaneously. We plan to hold at least 6GWh of internal capacity in South Korea. In Europe, however, we won't build from scratch. While shipping cells from Korea is an option, leveraging local idle capacity through strategic partners allows us to enter the European market far faster and more capital-efficiently.

Q: Since the battery foundry concept emerged, making prototype or R&D cells has always been possible. But when major players like Morrow or this French firm approach you for large-scale production, have domestic battery makers shown similar interest in handing over legacy fabs to JRX?

A: Domestic large enterprises operate under complex decision-making frameworks. For us to take over and operate an existing domestic site, managing onsite personnel, labor unions and corporate culture introduces significant operational friction, which slows down corporate decision-making. We presented aggressive proposals, but internal approvals stall. Overseas firms like Morrow or the French firm actively seek us out, whereas in Korea, we are the ones initiating proposals. While domestic executives acknowledge it as a great concept, execution remains complex.

Separately, taking over gigawatt-scale electrode lines incurs high operational costs if the site isn't fully decoupled from adjacent facilities. Domestic sites are tightly integrated. So while the strategic logic makes total sense —

Q: Even battery majors acknowledge the business logic.

A: They like the concept very much and actually initiated early discussions. We submitted formal proposals, but final approvals hit corporate administrative roadblocks. While discussions are currently paused, we believe domestic partnerships will eventually materialize.

Q: JRX's core business model is now thoroughly validated. With global cell makers seeking mass-production partnerships, what is your outlook on the broader battery market next year and beyond?

A: In the short term, our target markets are growing rapidly. ESS, robotics and drones are expanding fast. Although these represent high-mix, low-volume segments that seem small to major battery manufacturers, this market is being actively built by foundries like JRX and fabless developers. We project sustained, strong growth here.

Regarding the mainstream EV market, recovery will take time. While global market demand continues to grow, Korean battery makers face intense competition from Chinese suppliers. Meanwhile, market adoption in the U.S. has slowed down due to policy uncertainties surrounding the Trump administration. Industry consensus points toward a broader EV recovery around 2028-2029.

That is why our short-term strategy focuses heavily on non-EV sectors. These markets exhibit intense demand and short sales cycles. Unlike automotive projects that require two to three years of development and validation before generating revenue, non-EV projects translate into immediate orders. For the mainstream EV market, we are positioning ourselves around the AS replacement segment, which offers virtually uncapped growth potential. For instance, one major Korean battery maker we are negotiating with has roughly 100 legacy projects slated for AS production conversion.

Q: Because early-generation EVs are accumulating on roads.

A: Exactly. LG Energy Solution first supplied automotive batteries for the Chevy Volt back in 2008 — nearly 20 years ago. As vehicles age and models are discontinued, maintaining replacement cell supply becomes a major operational burden for automakers and battery firms. That is why our foundry outlook remains exceptionally bright.

Q: To wrap up, what are your key operational targets for next year, and what is your timeline for the upcoming IPO?

A: This year marked the successful validation of our foundry model. Next year will focus on scaling up that model — expanding market share and revenue.

Our 2026 revenue target is roughly 32 billion won, where revenue was strictly capped by physical capacity. Now, electrode capacity is expanding from 0.5GWh to 2GWh. We are also upgrading our cell assembly capacity from 100,000 cells to 300,000-400,000 cells by February next year. As production capacity quadruples, revenue will naturally scale.

Q: Mathematically, scaling capacity from 0.5GWh to 2GWh represents a fourfold potential increase.

A: At full utilization, that capacity translates to roughly 200 billion won in potential revenue. For next year, we are targeting over 100 billion won in revenue. This isn't an internal estimate — it is backed by major purchase orders (POs) scheduled for delivery by late this year from clients transitioning to mass production. Based on those POs, 100 billion won is fully achievable next year.

Q: What is your timeline for the IPO?

A: The IPO is a key milestone in our long-term growth strategy, primarily designed to raise public capital for major expansion opportunities. Beyond our existing Eumseong expansion, we are evaluating capital requirements for a 4GWh project and a 1-million-cell assembly line dedicated to drone battery production.

Starting next week, we are hosting a series of roadshows with Shinhan Securities for existing shareholders to disclose our IPO roadmap. Our target is to submit our formal IPO application in March 2027, with a target listing date in September 2027.

We selected lead underwriters back in November 2025 and have spent the past year establishing internal control systems, financial and accounting compliance, management infrastructure and corporate governance structures required for a public listing. While some internally questioned listing during a broader battery market downturn, our strategy is different: rather than relying on market hype, we intend to prove that our unique foundry model delivers stellar financial results even during industry downturns. We remain fully committed to submitting our application in March 2027 and listing in September 2027.

Intekplus will supply about 6 billion won worth of semiconductor and substrate inspection equipment overseas.

Intekplus said in a regulatory filing on Sept. 9 that it signed a contract with ACCUTECH to supply semiconductor and substrate inspection equipment. ACCUTECH is a semiconductor equipment distributor based in Shenzhen, China.

The contract is valued at about 6.08 billion won, equivalent to 6.78% of Intekplus’ 98.77 billion won in revenue last year.

Intekplus will supply two types of equipment from its iPIS series for semiconductor package and chip inspection and its iSIS series for semiconductor mid-end and substrate inspection. The specific models were not disclosed.

The iPIS lineup includes systems for final visual defect inspection of semiconductor packages, 3D dimensional inspection and 2D side inspection. The iSIS lineup includes systems for 3D inspection of microbumps on flip-chip substrates and 3D inspection of bumps and surface defects on flip-chip ball grid array (FC-BGA) substrates.

Payments will be made in four stages. Intekplus will receive 30% of the contract value as an advance payment and another 30% after equipment delivery. A further 30% will be paid after installation and commissioning are verified, with the remaining 10% due after final acceptance. The contract runs through Nov. 30 next year, with the final payment expected before then.

Intekplus recently won an order from a major Japanese substrate manufacturer for automated optical inspection (AOI) equipment used in glass-substrate research and development. The equipment will be delivered in the second half of this year. The number of systems and contract value were not disclosed.

Intekplus posted first-half revenue of 21.77 billion won and an operating loss of 7.03 billion won. Its first-half net loss widened from a year earlier to 7.54 billion won.

LG Innotek has unveiled an embedded flip-chip ball grid array (FC-BGA) substrate with silicon capacitors integrated inside the package substrate for artificial intelligence chips. The company is also developing technology to embed active devices and bridges. Demand for embedded substrates, particularly FC-BGAs, is growing as higher-performance AI chips require more stable power delivery.

LG Innotek displayed FC-BGA samples with embedded silicon capacitors at KPCA Show 2026, held at Songdo Convensia in Incheon on Sept. 9. The company showed two structures, with silicon capacitors embedded in either the substrate’s core layer or buildup layer.

The core-layer structure accommodates relatively thick, high-capacitance silicon capacitors. The buildup-layer structure embeds thinner silicon capacitors to reduce overall substrate thickness. LG Innotek said the location and specifications of the embedded silicon capacitors can be varied depending on the semiconductor and substrate design.

A silicon capacitor is a passive component made by forming dielectric and electrode layers on a silicon wafer. It stores power required by a semiconductor and supplies it instantaneously to suppress voltage fluctuations. Conventionally, capacitors are mounted on the substrate surface and deliver power to the chip through copper interconnects. Embedding capacitors inside the substrate shortens the power-delivery distance between the semiconductor and capacitor while freeing up mounting area on the substrate surface.

Silicon capacitors are particularly suited to stable power delivery at high frequencies because of their low equivalent series inductance (ESL). AI semiconductors operate at high speeds and can draw large amounts of power instantaneously. Lower ESL helps suppress voltage fluctuations.

Major silicon capacitor suppliers include Murata and Samsung Electro-Mechanics. TSMC also uses internally produced silicon-based deep trench capacitors (DTCs) in advanced packaging.

LG Innotek has been developing embedding technology for about a decade. It initially embedded multilayer ceramic capacitors (MLCCs) in flip-chip chip-scale packages (FC-CSPs). As the AI semiconductor market expanded, the company unveiled FC-BGA technology with embedded silicon capacitors in 2024 and began applying it to mass-produced products last year.

LG Innotek is also developing technology to embed active devices and bridges inside FC-BGAs, in addition to passive components such as silicon capacitors and inductors. Active devices directly control signals and power, while bridges connect multiple chips over short distances to support high-speed signal transmission. The goal is to expand the substrate’s role beyond simply connecting semiconductors to the mainboard.

“We have applied embedding technology for about 10 years, beginning with passive components inside FC-CSPs,” an LG Innotek official said. “It is attracting renewed attention as FC-BGAs become more advanced.”

“In the past, substrates simply served as a medium for transmitting signals, but they are now evolving to provide functionality themselves by incorporating silicon capacitors, inductors and other components,” the official said. “Moving components from the substrate surface to the inside gives customers greater design flexibility and enables thinner products.”

Inoplaztech plans to expand its annual production capacity for carbon nanotube (CNT) conductive additives to 100 tons. Conventional CNT additives in liquid dispersion form are difficult to use in dry electrode manufacturing, which does not use solvents. Inoplaztech aims to target the dry electrode market by using its proprietary plasma technology to disperse CNTs in powder form.

Inoplaztech CEO Lee Deok-yeon unveiled the company’s CNT conductive additive mass-production plan at the K-Battery Show 2026 Materials, Parts and Equipment Conference held Sept. 9 at KINTEX in Ilsan, Gyeonggi Province. Starting next year, the company plans to expand production facilities over three years to secure annual capacity of 100 tons. It plans to finance the investment through fundraising and programs including Scale-up TIPS, which supports the growth of small and medium-sized ventures.

CNTs are carbon materials used as conductive additives to improve electrical conductivity inside batteries. They can provide high conductivity in small amounts but tend to agglomerate, making them difficult to use as-is. A dispersion process is needed to separate CNT agglomerates and distribute them evenly throughout battery materials.

CNT conductive additives are currently produced mainly through a wet process in which CNTs are dispersed in a solvent using mechanical force. Dispersants are also used to prevent reagglomeration. The process can damage CNTs or leave residual dispersants that reduce conductivity. CNTs can also agglomerate again after the solvent is removed.

Inoplaztech disperses CNTs using plasma without solvents or dispersants. The process separates CNT agglomerates while treating their surfaces to suppress reagglomeration. Because the CNTs can be used as a powder rather than a liquid dispersion, they can be applied to dry electrode manufacturing.

Inoplaztech has been validating the applicability of its dry-dispersed CNTs with battery companies and others. It will now shift its focus from technology validation to securing mass-production capacity. The company expects annual sales of 20 billion to 30 billion won from its CNT conductive additive business once it reaches annual production capacity of 100 tons.

The company is also commercializing technology to improve CNT conductivity. The process uses plasma to attach nitrogen to the surface of multi-walled carbon nanotubes (MWCNTs), increasing their conductivity. Inoplaztech plans to begin selling nitrogen-functionalized MWCNTs as reagents by the end of this year. Such products remain relatively uncommon, with only three to four suppliers worldwide.

The goal is to improve the performance of relatively inexpensive MWCNTs so they can partially replace more expensive single-walled carbon nanotubes (SWCNTs). “Our long-term goal is to provide SWCNT-level performance at MWCNT prices,” Lee said.

Inoplaztech is also applying its plasma technology to other nanopowders. Nanopowder coating is a potential next business. As battery materials such as lithium iron phosphate (LFP) become finer, technologies capable of uniformly coating even small particles are increasingly needed. The company is considering plasma-based surface coating of fine powders as its next business area.

The dry plasma technology can also be applied outside batteries. Potential applications include polyethylene (PE) for battery separators, alumina fillers used in semiconductor thermal management materials and CNT composites for electric-vehicle tires. The underlying principle is the same: modifying surface properties so different materials can be mixed more effectively.

Inoplaztech is a nanomaterials company founded in 2018 as a startup by researchers from Ulsan National Institute of Science and Technology (UNIST). It has developed plasma technology that disperses aggregation-prone nanomaterials such as CNTs and graphene without solvents.

Coway Life Solution is partnering with integrated home care provider Wellcare Station to expand care services for customers with dementia and long-term care needs.

Coway Life Solution said Sept. 9 it signed a memorandum of understanding with Wellcare Station to support home care services for senior customers who need care and their families.

Wellcare Station is an integrated home care provider where doctors, nurses and care workers collaborate to provide home nursing and home care services. It also offers in-home care services including hospital accompaniment and support for using South Korea’s long-term care family leave program.

Under the agreement, Coway Life Solution customers with family members who require continuous care due to dementia or other conditions can receive support for long-term care services.

The service covers checking eligibility for the long-term care family leave program, connecting customers with care workers and coordinating schedules. For customers without a long-term care grade, it also handles related administrative procedures, including preparing documents needed to apply for a grade and filing applications with the National Health Insurance Service.

Wellcare Station will operate a dedicated service channel for Coway Life Solution customers. Customers using home care services will also receive priority assignment of care workers on their preferred dates.

Coway Life Solution is also expanding dementia care and caregiving support for senior customers. In partnership with digital healthcare company Silvia Health, it provides cognitive function assessments and personalized cognitive training. It also has partnerships with caregiving and long-term care providers Caring and Caredoc.

“We will help Coway Life Solution customers access long-term care-related services more conveniently,” Wellcare Station CEO Kook Kyung-min said.

Samsung Electronics has begun developing its first artificial intelligence glasses equipped with a display. The device is designed to show color visual information across the front of the lens.

Samsung Electronics asked Samsung Display last month to develop a microdisplay for AI glasses, according to industry sources on Sept. 9. It provided requirements covering the product form factor, information display area, color and price.

Samsung is expected to evaluate the performance and cost of panels that meet its requirements, including OLED-on-Silicon (OLEDoS) and LED-on-Silicon (LEDoS). Industry officials expect the product to launch in the second half of next year or the first half of 2028, with the timing depending on panel development and product validation.

Samsung has considered an ultra-small panel measuring 0.2 inch or less that can fit inside an eyeglass frame. An optical system would magnify the panel image and deliver it to the user’s eye. The device is intended primarily to display simple information such as text, notifications, icons and navigation arrows. Samsung has yet to decide whether to use a binocular design with displays for both eyes or a monocular design for one eye.

Rather than a high-resolution display capable of finely reproducing colors and contrast, Samsung is pursuing limited gradation of about 4 bits. Gradation refers to the number of brightness levels. Four bits per color channel provides 16 levels, compared with 256 levels for 8 bits. The approach is expected to prioritize easy recognition of text and symbols rather than rich colors for photos and video.

Industry officials consider white OLEDoS a practical option in terms of cost and production yield. OLEDoS is an ultra-small display that forms organic light-emitting diodes (OLEDs) on a silicon substrate. The white approach combines white OLEDs with color filters to produce colors. RGB OLEDoS does not require color filters but is more difficult to manufacture at high yields and carries higher production costs.

LEDoS combines microscopic light-emitting diodes (LEDs) with a silicon substrate. Its strength is high brightness, but color reproduction and mass-production yields remain challenges. It also requires technologies such as integrating red, green and blue (RGB) elements or converting light from blue LEDs into red and green.

Another technical challenge is securing a wide field of view (FOV), which determines the angular range over which users can see the virtual display. Because Samsung wants information displayed across the front of the lens, the viewing angle must also be wide. The product must achieve this while meeting color and cost requirements.

Samsung is planning the AI glasses as a smartphone companion accessory, keeping the smartphone as the primary device while allowing users to view necessary information through the glasses.

Samsung’s first AI glasses, scheduled for release this year, operate without a display and focus on camera and voice functions. They were jointly developed with Google, Gentle Monster and Warby Parker, combining Samsung’s hardware engineering, Google’s AI technology and designs from the two eyewear brands.

Meta commercialized color-display AI glasses in September last year. Meta Ray-Ban Display uses a monocular design with a screen only in the right lens. The display has a resolution of 600×600 pixels and a 20-degree FOV. Its full-color screen can display messages, photos, maps and translated subtitles, with the display positioned to the side of the user’s field of vision to avoid obstructing the forward view. Samsung, by contrast, plans to display text and symbols across the front of the lens using limited gradation.

“Samsung plans to evaluate OLEDoS, LEDoS and other panels that meet its requirements before commercializing the product,” an industry official said. “Samsung has not specified a particular panel technology, but RGB OLEDoS presents significant cost and production-yield challenges. Realistically, white OLEDoS is the more likely option.”

SK keyfoundry said Sept. 9 it signed a long-term agreement to provide 130-nanometer foundry services to automotive semiconductor company Elmos Semiconductor.

SK keyfoundry recently held a signing ceremony in Frankfurt am Main, Germany, and will provide foundry capacity to Elmos through 2037. Elmos has used SK keyfoundry’s 350-nanometer process technology for the past 18 years.

“Elmos is an outstanding company in the automotive semiconductor market,” an SK keyfoundry official said. “Securing a long-term supply agreement is meaningful not only in terms of revenue but also for our technological growth.”

Wafer supply prices may be adjusted under the agreement. The 200-millimeter wafers supplied by SK keyfoundry are experiencing a global shortage as major chipmakers including TSMC and Samsung Electronics reduce 200-millimeter capacity and shift toward 300-millimeter production.

SK keyfoundry and Elmos will jointly develop the E550.01 Smart 4-Channel eFuse Controller. The chip controls the supply of power inside a vehicle across four separate paths and cuts power to protect electrical systems when abnormalities such as overcurrent occur. Mass production is targeted for 2027.

SK keyfoundry will also provide Elmos with deep trench isolation (DTI) capability. The company said the technology enables highly integrated, high-voltage designs with low substrate current.

“SK keyfoundry has been a trusted partner for many years,” Elmos Chief Executive Officer Arne Schneider said. “It will support our continued growth.”

“We will accelerate our automotive semiconductor supply, starting with Elmos,” SK keyfoundry CEO Lee Dong-jae said.

DB HiTek is in the final stages of preparing for mass production of its 8-inch silicon carbide (SiC) wafer foundry business.

DB HiTek said Sept. 9 that it had completed reliability verification of a 1,200-volt SiC MOSFET. The company plans to provide customers with its third-generation process design kit (PDK) in November as it moves to launch its foundry business.

The third-generation PDK targets specific on-resistance of 2.3 mΩ·cm² at an 18-volt gate voltage, down from 2.5 mΩ·cm² for the second generation. It is also designed to allow the device to withstand a sudden short circuit for 2.5 microseconds without damage. DB HiTek said the third-generation PDK could shorten semiconductor design time by more than a year.

“We have secured the core technology and mass-production foundation for what will be the world’s first 8-inch SiC foundry service,” DB HiTek said. “We will prepare for mass production as planned and move to establish an early lead in the market.”

Mass production is targeted for the first half of next year. Rival SK keyfoundry is also expected to begin mass production of 8-inch SiC foundry products in the first half of next year. SiC foundry operations require production lines to be developed and established jointly with customers, meaning both companies have effectively secured customers ahead of mass production. Samsung Electronics also has an 8-inch SiC production line, but does not offer it to external foundry customers.

The SiC semiconductor market currently centers on 150-millimeter wafers, while mass production using 200-millimeter wafers remains rare globally. SiC semiconductors use specially processed wafers that differ from conventional silicon (Si) wafers. Due to the technology’s lower maturity, production yields for 200-millimeter SiC wafers are lower than for 150-millimeter wafers, resulting in lower output.

DB HiTek will operate its SiC foundry at its Sangwoo Campus in Eumseong, North Chungcheong Province, converting the current development line into a mass-production line. Initial capacity will be relatively small at 800 wafers a month, about 1.3% of the Sangwoo facility’s total monthly capacity of 63,000 wafers. The company has established an integrated process covering design, manufacturing, characterization, reliability verification and epiwafer supply.

For now, DB HiTek is operating the development line primarily for customers it has already secured. It plans to add customers after establishing the mass-production system and gradually expand capacity in line with market demand.

Battery diagnostics company MONA is developing an artificial intelligence foundation model using data from 10 million battery cells. The goal is to create a universal model that can apply diagnostic capabilities learned from one type of battery to different batteries. The company also plans to expand diagnostics to electric vehicles, energy storage systems (ESS) and second-life batteries.

"We plan to secure data from 10 million battery cells and develop a universal diagnostic model that can be applied across different batteries," MONA CEO Jang Ikhwang said at the K-Battery Show 2026 Materials, Parts and Equipment Conference held Sept. 9 at KINTEX in Ilsan, Gyeonggi Province. MONA has so far collected data from about 8 million battery cells and 1,000 battery packs.

Current battery diagnostic AI models struggle to maintain accuracy when applied to batteries different from those used for training. Materials, designs and production conditions vary by battery, meaning the same abnormality can appear differently in measurement data. MONA is training on large datasets collected from various batteries to develop a universal model capable of determining quality and detecting abnormalities across different products.

MONA's ability to collect large datasets is based on battery diagnostic technology using Electrochemical Impedance Spectroscopy (EIS). EIS applies alternating-current signals to a battery and measures how impedance changes at different frequencies.

The technology can identify latent internal cell defects and quality variations that are difficult to detect through conventional voltage and visual inspections. Even when a battery has normal voltage at shipment, quality problems can emerge during transportation and storage due to differences in factors such as self-discharge.

Applying EIS to mass-production lines requires faster inspection because directly measuring multiple frequency ranges takes time. MONA developed a high-speed diagnostic technology that measures only part of the frequency range and uses generative AI to reconstruct the remaining impedance data. The company has reduced inspection time to about one second per cell, allowing it to accumulate large datasets on mass-production lines.

Fast measurement alone is not enough for accurate quality assessment. Impedance is sensitive to temperature, meaning readings from the same battery can vary depending on the measurement environment. MONA corrects variations caused by temperature and equipment conditions and establishes a normal range based on the statistical distribution of accumulated data. Batteries outside that range are automatically flagged as potential defects.

MONA has also confirmed correlations between its data and actual defects. The company identified batteries in accumulated incoming-inspection data whose internal resistance and charge-transfer characteristics fell outside the normal distribution. Computed tomography (CT) and cross-sectional analysis subsequently found physical defects, including uneven electrode thickness. The abnormalities detected solely from impedance data matched physical defects inside the cells.

MONA is extending data collected from mass-production sites beyond the manufacturing stage. It is expanding the technology to diagnose batteries installed in EVs and determine whether used batteries recovered from vehicles can be reused. The company can collect data throughout the battery lifecycle, from incoming inspection of cells, modules and packs to EV diagnostics and second-life battery reuse.

MONA is also testing the feasibility of using second-life batteries in ESS applications. The company is conducting pilot tests of ESS systems incorporating used battery modules and packs. It is also developing diagnostic solutions for next-generation batteries, including all-solid-state batteries and batteries using silicon anodes.

Battery thermal runaway can generate flames exceeding 1,200°C as well as pressurized debris. LG Chem has developed an engineering plastic called Super Flame Barrier (SFB) that withstands such extreme conditions to prevent battery fires from escalating into larger accidents.

Hyun Sung-jae, a team leader at LG Chem, presented SFB and other materials designed to prevent battery thermal runaway propagation, along with the company’s development results, at the K-Battery Show 2026 Materials, Parts and Equipment Conference held Sept. 9 at KINTEX in Ilsan, Gyeonggi Province.

Battery thermal runaway is particularly dangerous because it may not remain confined to a single cell. When a cell is damaged or develops an internal short circuit, its temperature and pressure rise rapidly. If the resulting heat raises the temperature of adjacent cells, it can trigger further thermal runaway in a chain reaction. A problem originating in a single cell can ultimately escalate into a fire or explosion involving the entire battery pack.

Preventing cascading thermal runaway requires blocking the transfer of flames and heat between cells. Inorganic materials such as ceramics, mica and aerogel are commonly used for this purpose. LG Chem instead turned to plastics. Engineering plastics can be injection-molded, making them easier to use in complex-shaped battery components. LG Chem added resistance to high-temperature flames while retaining that processability.

LG Chem developed a material called Flame Barrier for this purpose. When exposed to flames, the plastic forms multiple carbonized layers on its surface to prevent heat and flames from penetrating. In LG Chem tests that exposed 1-millimeter-thick samples to a 1,000°C flame, conventional flame-retardant plastics developed holes or melted within eight to 109 seconds. Flame Barrier maintained its shape for more than 600 seconds. The material has been used in electric-vehicle batteries from a European automaker and a global battery manufacturer and is also being supplied for energy storage systems (ESS).

LG Chem subsequently developed SFB with further improved heat resistance. SFB has also been spec-in approved by a global automaker and has entered the commercialization stage. The company unveiled SFB at InterBattery 2026 in March. The technology won an InterBattery Award in the Reliability, Safety and Sustainability category.

The key to SFB is a protective barrier that remains intact at temperatures exceeding 1,200°C. When exposed to flames, special flame retardants and inorganic materials form an aluminum phosphate-based ceramic layer. While the existing Flame Barrier slows flame penetration through carbonized layers, SFB uses a rigid ceramic layer to withstand both high-temperature flames and pressure. The barrier can also remain intact against debris generated during thermal runaway.

LG Chem also verified SFB’s performance in tests simulating actual thermal runaway conditions. In repeated tests involving high-temperature flames and high-pressure particles, SFB withstood temperatures of 1,050°C to 1,200°C more than 12 times. It showed no holes or melting after the tests.

In tests using actual cylindrical battery cells, flames did not spread to surrounding cells even when the center cell was made to explode. A tray using SFB also showed no deformation. LG Chem has expanded the SFB product lineup for use in module covers and housings, battery cell frames and busbar components.

“It is important to delay the propagation of thermal runaway to surrounding cells when a battery fire occurs,” Hyun said. “We will continue developing engineering plastic materials that can be applied to battery components.”

Samsung Electronics has expanded heart health management features based on its Galaxy Watch devices and Samsung Health app.

Samsung introduced features including blood pressure and electrocardiogram (ECG) measurement, a heart health score, vital signs monitoring and health data sharing among family members at a media briefing held Sept. 9 at the Taepyeong-ro Building near Seoul City Hall.

Samsung introduced an optical photoplethysmography (PPG) sensor with the Galaxy S5, Gear 2 and Gear Fit in 2014 and has since continued to expand its health management technologies. In 2020, its Samsung Health Monitor app for blood pressure and ECG measurement received medical device approval from South Korea’s Ministry of Food and Drug Safety.

In 2021, Samsung introduced the BioActive Sensor on the Galaxy Watch4, combining optical heart rate, ECG and bioelectrical impedance analysis (BIA) sensors. In 2023, its Irregular Heart Rhythm Notification feature, which alerts users to signs of possible atrial fibrillation, received clearance from the U.S. Food and Drug Administration (FDA) and South Korea’s Ministry of Food and Drug Safety.

The Galaxy Watch Ultra2 and Watch9 support a heart health score feature. It analyzes health data including sleep, activity levels, vascular stress and body composition to provide a score and health management guidance.

The devices also support a vital signs feature that tracks physical conditions during sleep. It measures five indicators — heart rate, heart rate variability, respiratory rate, skin temperature and blood oxygen saturation — and alerts users when changes differ from their usual baseline.

Samsung Health also includes health data sharing. With user consent, data including steps, exercise, sleep, weight, body composition, food intake, heart rate, stress, blood pressure, blood glucose, medication and vital signs can be selectively shared with family members or acquaintances.

Linking family profiles with Samsung accounts allows family members to view one another’s health records.

Samsung also disclosed real-world examples of the features in use. In Jordan last year, a user discovered atherosclerosis and underwent a procedure after receiving an irregular heart rhythm notification. In Brazil, a user who repeatedly received inconclusive ECG readings sought medical attention and addressed the risk of myocardial infarction, Samsung said.

“We are combining Galaxy Watch sensor technology with Samsung Health’s analytical capabilities to help users manage heart health risk factors,” said Choi Jong-min, executive vice president of the Digital Health Team at Samsung Electronics’ Mobile eXperience (MX) Business. “We will introduce smart healthcare services focused on prevention.”

Gaon Cable, a subsidiary of LS Cable & System, has entered Canada's public power grid market for the first time.

Gaon Cable said Sept. 9 it recently supplied medium-voltage (MV) cables to a major public utility in western Canada. It marks the company's first export of power cables to Canada.

Gaon Cable plans to use its entry into Canada as a springboard to expand orders in the North American public power grid market. Canada is expanding and modernizing its transmission and distribution networks to meet rising electricity demand driven by the growth of AI data centers and advanced manufacturing.

The company also expects its North American sales to increase. Gaon Cable is expanding local supply through its headquarters and U.S. manufacturing subsidiary LSCUS in North Carolina.

Gaon Cable expects North American sales to rise more than 50% to about $470 million this year from about $300 million last year.

"We will continue to strengthen our presence in the North American power infrastructure market spanning the U.S. and Canada," Gaon Cable CEO Chung Hyun said.

U.S. semiconductor equipment maker Axcelis Technologies said Sept. 8 it plans to build a new equipment manufacturing facility in South Korea with an investment of about $35 million.

Robert Mahoney, Axcelis' Executive Vice President of Global Operations, recently participated in an investment declaration ceremony in Washington, D.C., with officials from South Korea's Ministry of Trade, Industry and Resources (MOTIR) and the Korea Trade-Investment Promotion Agency (KOTRA). The investment totals $35 million.

The new manufacturing facility will be built in Pyeongtaek with a total gross floor area of about 200,000 square feet. The investment is aimed at expanding ion implantation equipment production capacity to meet global demand.

Ion implanters are key front-end semiconductor manufacturing tools that precisely introduce ions into wafers to create desired electrical properties. Axcelis Korea established the Axcelis Asia Operations Center (AAOC) in Pyeongtaek in 2021 and began producing the equipment locally.

The new facility will be designed around a lean manufacturing environment aimed at eliminating waste and maximizing production efficiency. A groundbreaking ceremony is scheduled for Oct. 20, with production targeted to begin in the second half of 2028.

The facility will include Class 1,000 (ISO 6) and Class 10,000 (ISO 7) cleanrooms, a warehouse and a training center. A Class 1,000 cleanroom limits particles measuring 0.5 micrometers (㎛) to no more than 1,000 per cubic foot.

"Ion implanters are among the most important tools in front-end semiconductor manufacturing," Axcelis President and Chief Executive Officer Russell Low said. The investment will strengthen the company's global high-volume manufacturing infrastructure and expand the capabilities needed to serve customers worldwide, he said.

Qualcomm has announced a semiconductor partnership with Amazon for artificial intelligence data centers, tied to warrants and a potential deal worth up to $60 billion.

Qualcomm said Sept. 8 it began a product collaboration with Amazon to build next-generation AI data center infrastructure. The partnership aims to supply cost-efficient custom application-specific integrated circuits (ASICs) for large-scale AI inference workloads on Amazon Web Services (AWS). ASICs are dedicated chips optimized for specific computing tasks or services.

"As AI demand accelerates, data center infrastructure will require advances in both computing and connectivity to deliver greater performance with more efficiency," Qualcomm Chief Executive Officer Cristiano Amon said. "Qualcomm is pleased to work with AWS on customized silicon and connectivity solutions, bringing decades of leadership in advanced processing and power-efficient compute, to deliver breakthrough performance and enable the next generation of AI infrastructure."

The companies will also jointly design high-performance optical connectivity solutions to address rising data traffic and bandwidth requirements in data centers. The solutions will target transmission speeds of up to 1.6 terabits per second (Tbps), using Qualcomm's high-speed serializer/deserializer (SerDes) and optical digital signal processor (DSP) technologies.

SerDes chips convert slower parallel signals inside semiconductors into a single high-speed serial data stream for transmission. Optical DSPs precisely compensate for signal distortion and loss as optical data travels through fiber.

Qualcomm Chief Financial Officer Akash Palkhiwala said the Amazon collaboration could generate a potential revenue opportunity of as much as $60 billion. He added that Amazon's selection validates Qualcomm's data center technology, strategy and ability to supply products at scale.

Up to $60 Billion Over 10 Years

The potential revenue cited by Palkhiwala is linked to Qualcomm's issuance of warrants, which give Amazon the right to purchase shares at a specified price in the future. Qualcomm plans to vest the warrants in stages based on Amazon's purchases of its data center ASICs and optical connectivity solutions. Amazon's maximum contracted purchase amount is $60 billion, with the agreement running for 10 years.

Amazon has already secured 15% of the warrants after committing to purchase Qualcomm products at the initial stage of the agreement. The product volume is estimated at about $9 billion.

Qualcomm has also received purchase orders (POs) from Amazon and begun chip production. Revenue from the chips is expected to be recognized starting in December. That makes Qualcomm's previously announced fiscal 2027 data center revenue forecast of about $5 billion increasingly likely to be achieved.

Qualcomm also plans to expand its use of Amazon's AI platform to shorten semiconductor design cycles. The company will increase the use of services including Amazon Bedrock in electronic design automation (EDA) workloads.

EDA involves using software to design semiconductor circuits and verify their performance. Amazon Bedrock is a fully managed generative AI service that allows developers to use pretrained generative AI models in cloud environments for their specific workflows and applications.

Qualcomm's data center ASICs are part of its Dragonfly portfolio. In addition to custom silicon, Dragonfly includes the C1000 central processing unit (CPU), AI300 AI inference accelerator and connectivity products.

Taesung said Sept. 9 it is showcasing glass substrate production equipment technology at the International Semiconductor Substrate & Advanced Packaging Industry Exhibition, or KPCA Show 2026.

KPCA Show is a South Korean exhibition specializing in printed circuit boards (PCBs) and semiconductor packaging. The event is being held Sept. 9-11 at Songdo Convensia in Incheon. Taesung is presenting process technologies optimized for glass substrates, including etching, cleaning and plating, at its booth.

In particular, the company is showcasing key equipment for the Through Glass Via (TGV) process. TGVs are tiny holes formed in glass substrates to create electrical signal paths between chips and substrates. Taesung is displaying etching equipment that shapes and refines the holes, cleaning equipment that removes contaminants and plating equipment that coats the holes with metal to enable electrical conductivity.

Taesung has developed equipment for each stage of glass substrate processing based on technology accumulated in its existing PCB manufacturing equipment business. The company is targeting the market with mass-production solutions tailored to different manufacturing methods and process conditions for individual customers. Supply discussions with major customers are currently in their final stages, and Taesung expects to secure new orders within this month.

"Global demand for glass substrates is growing as AI semiconductors become more powerful and advanced packaging technologies develop," Taesung said. "We will actively showcase our differentiated process equipment technology at KPCA Show."

The growing AI and high-performance computing (HPC) markets are also increasing the importance of next-generation semiconductor packaging technologies. Glass substrates, a next-generation packaging material, have flatter surfaces than conventional substrates made of plastic or organic materials, making them better suited for denser, finer circuit patterns. They are also less susceptible to warpage and deformation caused by heat.

Simmtech is showcasing next-generation semiconductor substrate technologies, including Small Outline Compression Attached Memory Module (SOCAMM) for artificial intelligence servers, at KPCA Show 2026, held Sept. 9-11 at Songdo Convensia in Incheon.

Key products on display include SOCAMM, next-generation Low Power Compression Attached Memory Module (LPCAMM), enterprise solid-state drive (SSD) modules and Compute Express Link (CXL) memory modules.

Simmtech is the largest supplier of module printed circuit boards (PCBs) and memory semiconductor package substrates used in SOCAMM. Nvidia's Rubin CPU uses eight SOCAMMs, with each SOCAMM containing four low-power DRAM (LPDDR) packages. Simmtech supplies both SOCAMM module PCBs and LPDDR package substrates.

The company is also showcasing ultra-thin and ultra-fine substrate technologies, including Ultra-Thin Substrate, Hybrid Substrate, Small Via, Fine Bond Finger and Fine-Pitch/Dual-Size Small Outline Package (SOP) technologies.

Simmtech is displaying substrate technologies designed to address rising power consumption and heat generation in AI semiconductors. Bar Via, Chunk Via and Thick Cu Pattern technologies improve heat dissipation for high-power semiconductors. The company is also showcasing Rigid-Flex Substrate using Cu Post and Cavity technologies, as well as Wettable Flank substrates.

Simmtech is also presenting preliminary research on glass substrates, a technology that replaces conventional organic cores with glass. The technology can reduce warpage in large-area packages while facilitating ultra-fine wiring and improving signal integrity.

"The importance of substrate technology is greater than ever as the AI semiconductor market grows," a Simmtech official said. "Through this exhibition, we will share our next-generation semiconductor substrate technologies and future vision with customers and industry officials and strengthen our technology leadership in the global market."

AI Spera CEO Kang Byung-tak strongly criticized contractors' reliance on vibe coding following the hacking incident involving the "Moduui Changeop" startup support platform.

He said a growing number of contractors are using artificial intelligence-generated code without sufficient verification, creating new security risks.

Kang cited the Moduui Changeop hack at the AIDC Security Technology Workshop held Sept. 7 at Militopia Hotel by Marine in Seongnam, Gyeonggi Province. "The websites built by contractors are a mess beyond imagination," he said.

Moduui Changeop is a startup support program run by the Ministry of SMEs and Startups. In June, non-public information belonging to 5,000 successful applicants in the program's first cohort was leaked externally, including email addresses, evaluation comments and summaries of startup ideas.

A government investigation found that some of the platform's application programming interfaces (APIs) contained non-public information. The data leaked while the APIs were being collected externally through methods including web crawling. The data was encrypted, but the encryption key needed to decrypt it was also exposed.

An AI solutions company participating in the program obtained non-public email addresses through abnormal API calls and other means. Controversy grew after the company sent promotional emails to those addresses. It was later revealed that about a month before the incident, an external security team had discovered a similar API vulnerability and reported it to the Moduui Changeop operator.

"News coverage has largely focused on the contractor having carried out the hacking, but in my view both issues are serious," Kang said.

The "two issues" Kang referred to were the actions of the company that obtained the non-public information and the platform's own security flaws. He argued that the fundamental problem was that the platform had been designed to allow such access in the first place.

"The code written by the contractor looks to me like it was probably created through vibe coding," Kang said. "Security was not considered at all."

Vibe coding is a development method in which users describe desired functions in natural language and AI generates the necessary code and user interface (UI). "Vibe coding is certainly a whole new world, but in my view it is a ticking time bomb," Kang said. "I think it has increased contractor supply-chain risks by more than 100-fold."

"ChatGPT or Claude does a very good job when you use it for security checks," he said. "But after fixing the issues, some functions stop working. If you use vibe coding again to fix those functions, new vulnerabilities emerge." Most of those problems involve permissions, he added.

Kang outlined major security problems found in systems built by contractors, including cases where users are authenticated but authorization checks are not performed; databases are left directly accessible from outside without firewalls or access restrictions; and actual client data is uploaded unchanged to test and development servers without access control lists (ACLs).

Kang also disclosed examples of security problems involving companies and public institutions beyond Moduui Changeop.

One involved an AI chatbot that Kang identified as a system belonging to the Agency for Defense Development. "AI chatbot screens containing important defense-related data are exposed on the internet and can simply be accessed," he said. "It is not even a case of the ID and password being 'admin admin' — there is no authentication at all. It actually works when you submit prompts."

He also cited what he described as a system belonging to the Financial Security Institute. "Access tokens and refresh tokens are exposed on the site as JSON files," Kang said. "This does not really even require hacking." Access and refresh tokens function as credentials issued by a server after login. If exposed, they can allow access with the account's privileges without a password.

Kang also cited examples involving private companies. Referring to a system he described as an AI service from Starbucks, he said, "You can see all the content immediately without logging in, and you can submit prompts as well."

Regarding an AI system for credit assessment developed by Upstage, Kang said, "There is no login, so not only can you look through all this data and freely submit prompts, but you can also see the actual data inside." He said the issue was particularly serious because the system contained various types of data, including apartment purchase contracts.

Kang said the examples illustrate how the external attack surface that companies must manage is expanding significantly in the AI era. "When people talk about hacking AI, they mostly think of prompt injection or attacks on AI models," he said. "But rather than hacking the AI itself, targeting administrator pages is much easier to compromise and attack."

AI Spera is a cybersecurity company founded in 2017.

Attack Surface Management (ASM), whose importance Kang emphasized by disclosing security issues at multiple companies and institutions, is AI Spera's core business. Its flagship solution, Criminal IP, collects and analyzes internet-connected IP addresses and servers to identify externally exposed assets and security risks.

Korea Aerospace Industries (KAI) is showcasing the KF-21 and Manned-Unmanned Teaming (MUM-T) systems in Poland as it seeks to expand its presence in the European market.

KAI said Sept. 9 it is participating in the International Defence Industry Exhibition (MSPO 2026), being held in Kielce, Poland, from Sept. 8 to 11.

KAI displayed its Next-generation Air Combat System (NACS), linking the KF-21, unmanned combat aerial vehicles (UCAVs), multi-purpose unmanned aerial vehicles (AAPs) and low-Earth-orbit communications satellites. It is also showcasing a MUM-T system connecting the FA-50 with an AAP.

In the rotary-wing segment, KAI unveiled systems linking Air-Launched Effects (ALEs) with the Marine Attack Helicopter (MAH) and Light Armed Helicopter (LAH). KAI plans to offer the MAH, whose system development was completed in August 2026, to European countries seeking to replace their helicopter fleets.

KAI is seeking follow-on business opportunities for the FA-50 in Poland. It also plans to pursue new projects in European countries with fighter replacement demand, including Romania, Bulgaria and Slovenia.

Hanwha Ocean has been selected as the preferred bidder for the Royal Thai Navy's next-generation frigate project worth about 680 billion won.

Hanwha Ocean said Sept. 9 that it was named the preferred bidder for the Royal Thai Navy's next-generation frigate acquisition program and will begin detailed negotiations toward a final contract.

The Royal Thai Navy has set a budget of 16.73 billion baht, or about 683.3 billion won, for the construction of one frigate and integrated logistics support.

Hanwha Ocean will negotiate technical specifications, project scope and contract terms with the Royal Thai Navy. The final project value and contract terms will be determined based on the outcome of the negotiations.

Defense companies from Spain, Singapore and Türkiye competed for the project alongside Hanwha Ocean. The South Korean shipbuilder previously delivered the frigate "Bhumibol Adulyadej" to the Royal Thai Navy in 2018.

Hanwha Ocean plans to expand follow-on naval defense cooperation with Thailand and pursue additional opportunities in Southeast Asia's surface combatant market.

Meta and Panmnesia have proposed a next-generation artificial intelligence (AI) data center architecture designed to make an entire data center operate like a single chip. The approach could address latency problems that emerge as AI models grow larger.

Panmnesia said Sept. 8 that it and Meta introduced technology in a Nature Reviews journal that connects central processing units (CPUs), accelerators and memory through Compute Express Link (CXL) and scales the architecture across a data center. Researchers from Meta and Panmnesia co-authored the paper.

The approach uses CXL to reduce variations in latency between racks, allowing a data center to operate predictably like a single chip. CXL is a high-speed interface that connects different devices including CPUs, graphics processing units (GPUs) and memory.

Racks, the building blocks of AI infrastructure, exchange data rapidly within each rack using high-speed link technologies. Standard interfaces such as Ethernet and InfiniBand are used to connect one rack to another.

The problem is that as AI models expand to trillions of parameters, latency with conventional interfaces increases sharply. AI companies have increased the number of CPUs and GPUs to train on massive datasets, inevitably resulting in higher latency.

Meta and Panmnesia addressed latency variation by connecting racks through CXL. They added three types of hardware: high-fanout switches, fabric controllers and link acceleration units. They also proposed optical CXL to reduce the physical distance over which signals must travel.

The number of accelerators that a single CPU can manage increased from two to 16, while up to 960 connected domains can be managed. Latency was reduced by as much as 90%. Panmnesia said the CXL-based technology demonstrated that models larger than today's AI models can be trained without interruption.

"CXL presents a direction for next-generation AI infrastructure in which a data center operates as a single computing system," Panmnesia CEO Jung Myoungsoo said.