LG AI Research has embarked on developing an artificial intelligence service that predicts stock price trends for South Korean listed companies such as Samsung Electronics (005930.KS) and SK Hynix (000660.KS). Following its entry into the U.S. market, the research arm is now expanding its predictive scope to the South Korean market, accelerating the commercialization of AI-based financial solutions.<\/p>\n
According to IT industry sources on the 6th, LG AI Research is collaborating with Koscom to develop an ‘AI Powered Equity Forecast Score (AEFS)’ dedicated to the domestic stock market. AEFS is a service where AI analyzes corporate time-series data to predict how stock prices will fluctuate over the subsequent four weeks.<\/p>\n
LG AI Research previously launched a U.S. stock market-specific AEFS in September last year in partnership with the London Stock Exchange Group (LSEG). That version was developed by training LG AI Research’s hyperscale AI model ‘EXAONE’ on LSEG’s financial data. For this South Korean version, the plan is to utilize listed company data held by Koscom, a subsidiary of the Korea Exchange.<\/p>\n
LG AI Research’s stock prediction AI has already undergone years of validation. A prototype of the stock prediction AI model was developed in 2022 and secured first place among 160 participating companies at the ‘M6 Time Series Forecasting Competition’ held that same year. The following year, the AI-powered exchange-traded fund (ETF) ‘LQAI’ was launched in partnership with U.S. asset manager Qraft Technologies. The product analyzes the 500 largest U.S. listed companies and invests in the 100 with the highest potential for share price appreciation. From the beginning of this year through the 2nd, LQAI recorded a return of 17.5%, outpacing the Nasdaq’s return of 11.1% over the same period.<\/p>\n
The greatest strength of AEFS is that the AI can explain the rationale and process behind its predictions. LG AI Research built a cross-validation system utilizing four AI agents and trained the AI exclusively on data held by highly authoritative institutions such as the LSEG and the U.S. Securities and Exchange Commission (SEC). This approach, the institute explains, has reduced the ‘hallucination’ phenomenon that can occur with general-purpose AI chatbots like ChatGPT or Gemini.<\/p>\n
“Because AI chatbots learn from posts on internet communities, a bias occurs where the wishes of existing investors are reflected in predictions,” an LG AI Research official explained. “We developed a reliable AI model by training it only on curated data.”<\/p>\n
LG Group plans to apply this AI prediction model to its affiliates as well. For instance, data from battery manufacturer LG Energy Solution (373220.KS) could be used to predict the market price of lithium, a key battery raw material. Within South Korea’s AI industry, LG AI Research is evaluated as the first among companies participating in national AI projects to have discovered a revenue-generating business using AI. This contrasts with SK Telecom (017670.KS), which has yet to decide on monetizing its AI service ‘A.’, and startups like Upstage and Motif, which are still developing their AI models.<\/p>\n
Meanwhile, the government has designated ‘Physical AI’ as the new battleground for AI competition and identified the next three years as a golden window. Unlike generative AI, physical AI’s core lies in the ‘action data’ accumulated as robots perform real-world tasks. The assessment is that South Korea can seize leadership based on its manufacturing competitiveness, as this is a field where sufficient data has not yet been accumulated globally.<\/p>\n
The Ministry of Science and ICT is simultaneously pursuing technology development, field verification, and institutional improvements with manufacturing data, world models, and AI semiconductors as its three pillars. In particular, because there are limits to manually collecting action data from actual sites, the ministry views the ‘world model’\u2014where AI learns the physical laws of the real world to implement a virtual environment\u2014and the ‘synthetic data’ generated through it as key alternatives.<\/p>\n
Kim Wook, Physical AI PM at the Ministry of Science and ICT, explained the world model concept: “Just as a person mentally runs through multiple scenarios\u2014whether to take the subway or the bus to a destination\u2014and makes the optimal choice, the AI also rapidly runs through countless scenarios within the world model and learns the best course of action.” The government anticipates that 80-90% of physical AI training data will be replaced by synthetic data going forward.<\/p>\n
However, the government’s stance is that even world models must not rely on foreign technology. Park Tae-wan, ICT Industry Policy Officer at the Ministry of Science and ICT, emphasized, “If we use ‘Cosmos,’ the world model developed by Nvidia, all the core data from our manufacturing sites will leak out. Advanced conglomerates like Samsung and SK Hynix are extremely sensitive, insisting that their data must never be leaked. We absolutely need a domestic world model to seize the core of physical AI.”<\/p>\n
The government’s confidence in physical AI stems from South Korea’s manufacturing sites. As one of the few countries with a comprehensive industrial base spanning semiconductors, automobiles, shipbuilding, and home appliances, South Korea can leverage its strengths in the manufacturing sector, where physical AI is expected to create value first. In fact, a physical AI pilot project conducted last year at eight factories in South Gyeongsang Province and three in North Jeolla Province reportedly resulted in productivity improvements of over 20%.<\/p>\n
The physical AI-based unmanned factory platform ‘KAIROS,’ developed by the Korea Advanced Institute of Science and Technology (KAIST), is also drawing attention. This platform, where collaborative robots, logistics robots, and humanoid robots move organically, is implemented with 100% domestic technology. When a single manager communicates and gives instructions to the AI via voice, the AI autonomously diagnoses and resolves the cause of anomalies.<\/p>\n
Song Chang-jong, head of the Device AX Innovation Team at the Ministry of Science and ICT, assessed that “there are no missing companies in the domestic physical AI full-stack domain.” This means that competitive companies are evenly positioned domestically in each core area constituting the full stack, including AI models, semiconductors, robots, and simulators. However, the fact that these companies have not yet coalesced into a single ecosystem is cited as a challenge. The Ministry of Science and ICT launched the second phase of the ‘Physical AI Alliance’ last month, focusing on securing full-stack technology and strengthening execution capabilities.<\/p>\n
Deputy Prime Minister and Minister of Science and ICT Lee Jong-ho stated at a national report event on the 29th of last month, “To properly foster the physical AI industry, it is crucial to secure our own technological capabilities without being dependent on the technological prowess of the U.S. and China. Given that South Korea is equipped with high-quality data and optimal sites, becoming the top power in physical AI is by no means just a lofty aspiration.”<\/p>\n","protected":false},"excerpt":{"rendered":"LG AI Research has embarked on developing an artificial intelligence service that predicts stock price trends for South…\n","protected":false},"author":2,"featured_media":76622,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[25],"tags":[41647,24654,41651,6944,10289,315,9196,314,41649,41648,1042,2295,1223,41650,276,275,16894],"class_list":["post-76621","post","type-post","status-publish","format-standard","has-post-thumbnail","category-lg","tag-aefs","tag-exaone","tag-kairos","tag-kaist","tag-koscom","tag-lg","tag-lg-ai-research","tag-lg-group","tag-lqai","tag-lseg","tag-ministry-of-science-and-ict","tag-nvidia","tag-physical-ai","tag-qraft-technologies","tag-samsung-electronics","tag-sk-hynix","tag-world-model"],"_links":{"self":[{"href":"https:\/\/www.europesays.com\/korea\/wp-json\/wp\/v2\/posts\/76621","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.europesays.com\/korea\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.europesays.com\/korea\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/korea\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/korea\/wp-json\/wp\/v2\/comments?post=76621"}],"version-history":[{"count":0,"href":"https:\/\/www.europesays.com\/korea\/wp-json\/wp\/v2\/posts\/76621\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/korea\/wp-json\/wp\/v2\/media\/76622"}],"wp:attachment":[{"href":"https:\/\/www.europesays.com\/korea\/wp-json\/wp\/v2\/media?parent=76621"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.europesays.com\/korea\/wp-json\/wp\/v2\/categories?post=76621"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.europesays.com\/korea\/wp-json\/wp\/v2\/tags?post=76621"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}