{"id":737,"date":"2026-07-13T14:50:35","date_gmt":"2026-07-13T05:50:35","guid":{"rendered":"https:\/\/blog.hancom.com\/artificial-intelligence-a-new-era-korea-ec-percent-a0-percent-95-percent-ec-percent-b1-percent-85-percent-ea-percent-b3-percent-bc-percent-ec-percent-82-percent-ac-percent-eb-percent-a1-percent-80-per\/"},"modified":"2026-07-13T15:19:14","modified_gmt":"2026-07-13T06:19:14","slug":"what-is-sovereign-ai","status":"publish","type":"post","link":"https:\/\/blog.hancom.com\/en\/what-is-sovereign-ai\/","title":{"rendered":"What Is Sovereign AI? From the Concept to Domestic and Global Policies and Cases"},"content":{"rendered":"\n<pre class=\"wp-block-preformatted\">\ud83e\uddd1\u200d\ud83d\udcbb Summary<br\/><br\/>This article first looks at the concept of Sovereign AI\u2014securing control over data, computing infrastructure, AI models and software, and governance and ethical frameworks without excessive dependence on the technologies of specific countries or companies\u2014and its four key components. It then introduces the Korean government\u2019s independent AI foundation model project, and Hancom\u2019s role as a core partner in the LG AI Research consortium in expanding the foundation for AI use of public and enterprise data based on its electronic document and document AI technologies. Finally, it examines how the push to secure AI sovereignty is spreading through Sovereign AI strategies and cases across countries including Japan, Singapore, the EU, and Canada.  <br\/><\/pre>\n\n<p class=\"wp-block-paragraph\">&#8220;Every country needs Sovereign AI.&#8221;<\/p>\n\n<p class=\"wp-block-paragraph\">These were the words of NVIDIA CEO Jensen Huang at the World Governments Summit (WGS) in Dubai in February 2024. That day, he told attendees that \u201cevery country needs to build and own its own AI.\u201d After this remark, Sovereign AI began drawing even more attention in global AI policy and across industry. According to the <a href=\"https:\/\/www.cambridge.org\/core\/journals\/natural-language-processing\/article\/sovereign-ai-in-2025\/C51560626AF518BDF280891C406E9553\" target=\"_blank\" rel=\"noopener\">OECD AI Policy Observatory<\/a>, more than 70 countries and regions have announced official AI strategies or plans, and many are strengthening Sovereign AI strategies to secure sovereignty over data, infrastructure, and models.   <\/p>\n\n<p class=\"wp-block-paragraph\"><strong>So what exactly is Sovereign AI, and how are countries around the world\u2014including Korea\u2014responding to this trend? <\/strong><\/p>\n\n<p class=\"wp-block-paragraph\">In this article, we explore everything from the concept of Sovereign AI to the Korean government\u2019s independent AI foundation model project and Sovereign AI strategies around the world.<\/p>\n\n<div style=\"height:10px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n<h2 class=\"wp-block-heading\"><strong><strong>What Is Sovereign AI?<\/strong><\/strong><\/h2>\n\n<pre class=\"wp-block-preformatted\">Sovereign AI refers to the capability to secure control over the data, computing infrastructure, AI models and software, and governance and ethical frameworks needed to develop, operate, and use AI\u2014without excessive dependence on the technologies of specific countries or companies.<\/pre>\n\n<div style=\"height:10px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n<div class=\"wp-block-image\">\n<figure class=\"aligncenter size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"561\" src=\"https:\/\/blog.hancom.com\/wp-content\/uploads\/2026\/07\/AX-1-1024x561.png\" alt=\"Image explaining the concept of Sovereign AI (Sovereign AI)\" class=\"wp-image-735\" srcset=\"https:\/\/blog.hancom.com\/wp-content\/uploads\/2026\/07\/AX-1-1024x561.png 1024w, https:\/\/blog.hancom.com\/wp-content\/uploads\/2026\/07\/AX-1-300x164.png 300w, https:\/\/blog.hancom.com\/wp-content\/uploads\/2026\/07\/AX-1-768x421.png 768w, https:\/\/blog.hancom.com\/wp-content\/uploads\/2026\/07\/AX-1.png 1283w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n<\/div>\n<div style=\"height:10px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n<h3 class=\"wp-block-heading\"><strong>The Four Components of Sovereign AI<\/strong><\/h3>\n\n<p class=\"wp-block-paragraph\">Sovereign AI is not a single technology, but a concept focused on securing decision-making authority and control across the core elements that make up AI. It is generally explained around the following four elements. <\/p>\n\n<ul class=\"wp-block-list\">\n<li><strong>Data sovereignty<\/strong>: Control over generating, storing, and using a country\u2019s data<\/li>\n\n\n\n<li><strong>Computing sovereignty<\/strong>: Securing the infrastructure resources needed for AI training and inference<\/li>\n\n\n\n<li><strong>Model and software sovereignty<\/strong>: The ability to design, deploy, and manage one\u2019s own AI models<\/li>\n\n\n\n<li><strong>Governance and ethical sovereignty<\/strong>: The authority to set AI safety and ethical standards suited to a country\u2019s circumstances<\/li>\n<\/ul>\n\n<div style=\"height:10px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n<h3 class=\"wp-block-heading\"><strong>Three Reasons Sovereign AI Is Needed <\/strong><\/h3>\n\n<p class=\"wp-block-paragraph\">The need for Sovereign AI can largely be summarized into three areas: national security, national competitiveness, and preserving identity. <\/p>\n\n<ul class=\"wp-block-list\">\n<li><strong>Strengthening national security: <\/strong>If sensitive national data\u2014such as defense, healthcare, and public administration\u2014relies on foreign big tech cloud or AI platforms, risks such as data leakage and external control may arise. <\/li>\n\n\n\n<li><strong>Securing national competitiveness:<\/strong> Customized AI models that reflect a country\u2019s industrial characteristics and high-quality data create greater value and productivity than general-purpose AI models.<\/li>\n\n\n\n<li><strong>Preserving identity and culture: <\/strong>Relying only on big tech models can distort linguistic, historical, and cultural values, making it necessary to train on data that reflects a country\u2019s language and culture. <\/li>\n<\/ul>\n\n<div style=\"height:10px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n<h2 class=\"wp-block-heading\"><strong>What Is the Independent AI Foundation Model Project?<\/strong><\/h2>\n\n<pre class=\"wp-block-preformatted\">The independent AI foundation model project is a national initiative \ucd94\uc9c4 by the Republic of Korea in response to the need for Sovereign AI, aiming to secure AI sovereignty and achieve technological self-reliance. It is part of the policy to leap into the \u201ctop three AI powers,\u201d and is often referred to as the \u201cNational Team AI Project.\u201d <\/pre>\n\n<div style=\"height:10px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n<h3 class=\"wp-block-heading\"><strong>Why an Independent AI Foundation Model Is Needed<\/strong><\/h3>\n\n<p class=\"wp-block-paragraph\">An AI foundation model is a base model trained on broad data that can be used for a wide range of AI tasks. Many AI services\u2014such as chatbots, document writing, image understanding, search, and \uc5c5\ubb34 automation\u2014are built on these foundation models. Because multiple AI services can be developed from a single foundation model, securing and leveraging the right foundation model has emerged as a key factor in AI competitiveness.   <br\/><br\/>Against this backdrop, the Korean government judged that as global AI competition spreads around hyperscale models and dependence on specific countries and companies deepens, the security, economic, and data sovereignty risks for public services and national AX initiatives increase. Accordingly, it is working to secure an \u201cindependent AI foundation model\u201d based on domestic technology to use as the base model for national AI transformation (AX), and to strengthen domestic AI competitiveness in line with the goals of Sovereign AI. <\/p>\n\n<div style=\"height:10px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n<h3 class=\"wp-block-heading\"><strong>Companies Participating in the National Team AI Project <\/strong><\/h3>\n\n<p class=\"wp-block-paragraph\">Through document review and presentation evaluations in August 2025, the independent AI foundation model project initially selected <a href=\"https:\/\/www.msit.go.kr\/bbs\/view.do?sCode=user&amp;mId=307&amp;mPid=208&amp;bbsSeqNo=94&amp;nttSeqNo=3186073\" target=\"_blank\" rel=\"noopener\">five elite teams<\/a> (consortia). Each consortium, led by domestic AI companies with participation from universities and research institutes, is \ucd94\uc9c4 the development of an independent AI foundation model.  <\/p>\n\n<ul class=\"wp-block-list\">\n<li><strong>NAVER Cloud<\/strong>: An omni foundation model that integrates understanding and generation across heterogeneous data such as images and documents<\/li>\n\n\n\n<li><strong>Upstage<\/strong>: A global frontier-level AI foundation model<\/li>\n\n\n\n<li><strong>SK Telecom<\/strong>: Next-generation transformation AI models and Korean-style AI services<\/li>\n\n\n\n<li><strong>NC AI<\/strong>: Large-scale language and multimodal AI foundation models<\/li>\n\n\n\n<li><strong>LG Management Development Institute AI Research<\/strong>: An AI foundation model combining specialization and general-purpose capabilities<\/li>\n<\/ul>\n\n<div style=\"height:10px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n<h3 class=\"wp-block-heading\"><strong>Hancom Joins the National Team AI Project as a Core Partner of LG AI Research <\/strong><\/h3>\n\n<p class=\"wp-block-paragraph\">To realize Sovereign AI, it is necessary not only to secure one\u2019s own AI models, but also to have foundational technologies that convert documents and data held by the public sector and enterprises into forms AI can use, and enable safe search and utilization.<\/p>\n\n<p class=\"wp-block-paragraph\">Based on 36 years of accumulated electronic document technology and experience applying AI in public and enterprise markets, Hancom is participating as a core partner in the LG AI Research consortium for the independent AI foundation model project. Through its products\u2014such as the AI-based document authoring tool \u201cHancom Assistant,\u201d the AI search solution \u201cHancom Pedia,\u201d and the AI training parsing solution \u201cHancom Data Loader\u201d\u2014Hancom supports AI-driven \uc5c5\ubb34 transformation in the public and enterprise sectors. Hancom has also secured a range of adoption cases in the public AI market that demands high security and technical capability, including the National Assembly Library, the Korea Coast Guard, Korea Western Power, and the National Assembly Secretariat. This experience plays an important role in underpinning key elements of Sovereign AI, including data sovereignty, security control, and the foundation for utilizing public and industrial data.   <br\/><br\/>In this way, through participation in the LG AI Research consortium, Hancom is expanding the foundation for using independent AI models and contributing to the spread of Korea\u2019s Sovereign AI ecosystem across B2B and B2G domains.<\/p>\n\n<div style=\"height:10px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n<h3 class=\"wp-block-heading\"><strong>National Team AI Project Progress (as of July 2026) <\/strong><\/h3>\n\n<p class=\"wp-block-paragraph\">Through document review, presentation evaluations, and stage evaluations, the government will narrow down participating teams on a six-month cycle and <strong>plans to select the final two teams by February 2027<\/strong>. <\/p>\n\n<p class=\"wp-block-paragraph\">After the initial selection of five elite teams in August 2025, NAVER Cloud and NC AI were eliminated in the first stage evaluation held in January 2026. LG AI Research received the highest score at 90.2 points and advanced to the second evaluation along with Upstage and SK Telecom. In February, Motif Technologies was additionally selected as an elite team, and the project is currently \uc9c4\ud589 under a four-team structure, with the second evaluation scheduled for early August.  <\/p>\n\n<div style=\"height:10px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n<h2 class=\"wp-block-heading\"><strong>Global Sovereign AI Strategies and Key Cases (Focused on Middle Powers) <\/strong><\/h2>\n\n<p class=\"wp-block-paragraph\">So what Sovereign AI strategies are middle powers with conditions similar to Korea pursuing? <\/p>\n\n<p class=\"wp-block-paragraph\">According to the National Security Strategy Institute (INSS) report <a href=\"https:\/\/www.inss.re.kr\/upload\/bbs\/BBSA05\/202511\/F20251121132200284.pdf\" target=\"_blank\" rel=\"noopener\">\u201cExploring a Korean-Style Sovereign AI National Strategy,\u201d<\/a> most middle powers outside the U.S. and China recognize resource constraints and are pursuing \u201cspecialized\u201d or \u201ccoalition and norms-based\u201d strategies. Rather than building every \uc601\uc5ed of AI, they focus on areas where they have strengths or join forces with like-minded countries.  <\/p>\n\n<h4 class=\"wp-block-heading\">Specialized \u2014 Focus on Areas of National Strength <\/h4>\n\n<ul class=\"wp-block-list\">\n<li><strong>Japan<\/strong>: By focusing on high-performance computing (HPC) and basic research infrastructure, Japan is internalizing national AI infrastructure by building the ultra-large AI supercomputing infrastructure \u201cABCI 3.0.\u201d<\/li>\n\n\n\n<li><strong>Singapore<\/strong>: Through the development of \u201cSea-Lion LLM,\u201d Singapore is seeking to secure language models optimized for Southeast Asian languages.<\/li>\n<\/ul>\n\n<h4 class=\"wp-block-heading\">Coalition and Norms-Based \u2014 Cooperation Among Multiple Countries <\/h4>\n\n<ul class=\"wp-block-list\">\n<li><strong>France and Germany (EU)<\/strong>: By establishing the \u201cGAIA-X framework,\u201d they have completed a technical and governance framework to ensure data interoperability, trust, and sovereignty.<\/li>\n\n\n\n<li><strong>African Union (AU)<\/strong>: By \ucd94\uc9c4 continent-wide data policies, the AU is promoting the sharing of training data among member states and building shared foundations.<\/li>\n<\/ul>\n\n<div style=\"height:10px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n<h2 class=\"wp-block-heading\"><strong>In the Era of Sovereign AI, What Matters More Than Technology Is \u201cAI Sovereignty\u201d <\/strong><\/h2>\n\n<p class=\"wp-block-paragraph\">AI competitiveness is no longer determined solely by having one outstanding AI model. AI sovereignty\u2014the ability to use data safely and develop and operate AI aligned with each country\u2019s environment and norms\u2014is emerging as a new source of competitiveness. <br\/><br\/>Amid this trend, Korea is also moving to secure Sovereign AI capabilities by \ucd94\uc9c4 the independent AI foundation model project. Hancom, too, is contributing to the expansion of the domestic Sovereign AI ecosystem by participating as a core partner in the LG AI Research consortium. In addition, as data sovereignty becomes more important in the era of Sovereign AI, Hancom recently declared a shift into a \u201cSovereign Agentic OS\u201d company. It is focusing on developing an AI agent operating system that connects and controls data, AI models, \uc5c5\ubb34 systems, and permission frameworks within a single secure environment.    <\/p>\n\n<p class=\"wp-block-paragraph\">Going forward, Sovereign AI will become an important standard not only in national policy but also in corporate AI strategies. Centered on the Sovereign Agentic OS, Hancom will build a trustworthy AI environment and continue to support the expansion of AI ecosystems at home and abroad and AI transformation (AX). <\/p>\n\n<div style=\"height:10px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n<h2 class=\"wp-block-heading\"><strong>Today\u2019s Summary <\/strong><strong>Q&amp;A<\/strong><\/h2>\n\n<h3 class=\"wp-block-heading\"><strong>Q1. What is Sovereign AI? <\/strong><\/h3>\n\n<p class=\"wp-block-paragraph\">Sovereign AI refers to the capability to control a country\u2019s own data and AI models without being dependent on the technologies of specific countries or companies. It consists of four elements\u2014data sovereignty, computing sovereignty, model and software sovereignty, and governance and ethical sovereignty\u2014and its importance is growing in terms of national security, national competitiveness, and preserving identity and culture. <\/p>\n\n<h3 class=\"wp-block-heading\"><strong>Q2. What policies is Korea pursuing for Sovereign AI?  <\/strong><\/h3>\n\n<p class=\"wp-block-paragraph\">A \ub300\ud45c Sovereign AI initiative of the Korean government is the \u201cindependent AI foundation model\u201d project. Based on the recognition that the more a country depends on AI models from specific countries or companies, the greater the risks to security, the economy, and data sovereignty, the government is \ucd94\uc9c4 this project with the goal of securing AI sovereignty and achieving technological self-reliance. Often called the \u201cNational Team AI Project,\u201d it began with five elite teams in August 2025 and, as of July 2026, four teams\u2014LG AI Research, SK Telecom, Upstage, and Motif Technologies\u2014are preparing for the second evaluation.  <\/p>\n\n<h3 class=\"wp-block-heading\"><strong>Q3. What role does Hancom play in Korea\u2019s Sovereign AI market?  <\/strong><\/h3>\n\n<p class=\"wp-block-paragraph\">Hancom is participating as a core partner in the LG AI Research consortium for the independent AI foundation model project. Through the AI-based document authoring tool \u201cHancom Assistant,\u201d the AI search solution \u201cHancom Pedia,\u201d and the AI training parsing solution \u201cHancom Data Loader,\u201d Hancom is helping expand the AI ecosystem in the B2B and B2G sectors and contributing to the spread of the domestic Sovereign AI ecosystem. Hancom has also recently declared a shift into a \u201cSovereign Agentic OS\u201d company and is focusing on developing an AI agent operating system that connects and controls data, AI models, \uc5c5\ubb34 systems, and permission frameworks within a single secure environment.   <\/p>\n\n<div style=\"height:10px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n<p class=\"wp-block-paragraph\"><strong>\ud83d\udcc4 References<\/strong><\/p>\n\n<ul class=\"wp-block-list\">\n<li><a href=\"https:\/\/www.cambridge.org\/core\/journals\/natural-language-processing\/article\/sovereign-ai-in-2025\/C51560626AF518BDF280891C406E9553\" target=\"_blank\" rel=\"noopener\">Robert Dale<\/a>, \u201cSovereign AI in 2025\u201d, <em>Natural Language Processing<\/em>, 2025. <\/li>\n\n\n\n<li><a href=\"https:\/\/www.inss.re.kr\/upload\/bbs\/BBSA05\/202511\/F20251121132200284.pdf\" target=\"_blank\" rel=\"noopener\">National Security Strategy Institute<\/a>, \u201cExploring a Korean-Style Sovereign AI National Strategy\u201d, 2025<\/li>\n\n\n\n<li><a href=\"https:\/\/www.msit.go.kr\/bbs\/view.do?sCode=user&amp;mId=307&amp;mPid=208&amp;bbsSeqNo=94&amp;nttSeqNo=3186073\" target=\"_blank\" rel=\"noopener\">Ministry of Science and ICT<\/a>, \u201cLaunch Ceremony Held for the Independent Artificial Intelligence Base Model (AI Foundation Model) Program (Project)\u201d, 2025<\/li>\n<\/ul>\n\n<p class=\"wp-block-paragraph\"><strong>\ud83d\udc40 Recommended Content<\/strong><\/p>\n\n<ul class=\"wp-block-list\">\n<li><a href=\"https:\/\/hai-hancom.stibee.com\/p\/98\" target=\"_blank\" rel=\"noopener\">[HAI Newsletter]<\/a> \ud83d\udc4b Korea\u2019s AI ship sets sail with \ud798 <\/li>\n\n\n\n<li><a href=\"https:\/\/www.hancom.com\/news\/article\/detail\/12952?type=news&amp;artclClss=&amp;searchText=%EB%8F%85%EC%9E%90\" target=\"_blank\" rel=\"noopener\">[Hancom Official Press Release]<\/a> Hancom boards the \u201cNational Team AI\u201d project\u2026 35 years of document technology to drive the success of \u201cK-EXAONE\u201d<\/li>\n\n\n\n<li><a href=\"https:\/\/blog.hancom.com\/en\/kpaix-2026-hancom-ai\/\">[Hancom Blog] <\/a>[On-site Sketch] Public AI and the Sovereign Agentic OS showcased by Hancom \u2014 2026 Public AI Expo (KPAIX 2026)<\/li>\n<\/ul>\n\n<div style=\"height:30px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n","protected":false},"excerpt":{"rendered":"<p>With more than 70 countries entering the Sovereign AI race, what strategies is Korea pursuing? Explore the concept of Sovereign AI and its four components, the independent AI foundation model project, global strategies, and Hancom\u2019s latest moves\u2014all in one place. <\/p>\n","protected":false},"author":2,"featured_media":738,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[26,10],"tags":[],"class_list":["post-737","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-ax-insight","category-10"],"_links":{"self":[{"href":"https:\/\/blog.hancom.com\/en\/wp-json\/wp\/v2\/posts\/737","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/blog.hancom.com\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/blog.hancom.com\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/blog.hancom.com\/en\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/blog.hancom.com\/en\/wp-json\/wp\/v2\/comments?post=737"}],"version-history":[{"count":2,"href":"https:\/\/blog.hancom.com\/en\/wp-json\/wp\/v2\/posts\/737\/revisions"}],"predecessor-version":[{"id":744,"href":"https:\/\/blog.hancom.com\/en\/wp-json\/wp\/v2\/posts\/737\/revisions\/744"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/blog.hancom.com\/en\/wp-json\/wp\/v2\/media\/738"}],"wp:attachment":[{"href":"https:\/\/blog.hancom.com\/en\/wp-json\/wp\/v2\/media?parent=737"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/blog.hancom.com\/en\/wp-json\/wp\/v2\/categories?post=737"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/blog.hancom.com\/en\/wp-json\/wp\/v2\/tags?post=737"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}