Ha Jung-woo: Architect of South Korea’s Sovereign AI Future

Introduction

In June 2025, Dr. Ha Jung-woo, a respected computer scientist and technocrat, was appointed as South Korea's first Senior Presidential Secretary for AI Future Planning. This unprecedented move signaled a paradigm shift for the country’s AI strategy, reflecting both the urgency and the ambition to leap into the ranks of the world's top three AI powerhouses. Dr. Ha’s trajectory—from academic prodigy to research leader at Naver, to national public servant spearheading sovereign AI—embodies South Korea’s evolving approach to technology governance. This comprehensive report examines his biography, academic credentials, research and leadership at Naver (especially the landmark HyperCLOVA X project), his multi-faceted contributions to government policy, ethical debates, civic engagement, and his current role charting the country’s high-profile AI investment and national strategy.

Biography and Early Education

Ha Jung-woo was born in Busan in 1977, a period when South Korea was rapidly industrializing but had yet to emerge as a global technology leader. He graduated from Gudeok High School in Busan before entering Seoul National University (SNU), Korea’s premier academic institution, where he completed a bachelor’s degree in Computer Engineering. Committed to advancing the frontier of artificial intelligence, he pursued and obtained a master's and then a Ph.D. from SNU’s Department of Electrical and Computer Engineering. His doctoral studies, guided by luminaries like Byoung-Tak Zhang, focused on machine learning, computer vision, and natural language processing—the bedrock fields for modern-day AI systems.

This period was foundational not only academically, but also in signaling his future orientation towards large-scale, practical impact. As part of a generation of Korean computer scientists who matured in the post-AlphaGo AI boom, Ha’s educational trajectory primed him for roles at the nexus of academia, industry, and ultimately, national policy.

Academic Credentials and Research Publications

Ha Jung-woo’s academic credentials are distinguished by both depth and breadth in AI. After undergraduate and graduate studies at SNU, where he specialized in electrical and computer engineering, he completed a landmark dissertation early in the deep learning era, with publications analyzing topics such as video understanding using animation data (notably “Pororo” video), years before the global surge in modern transformer-based architectures. His work straddled areas like continual learning, generative models, deep learning, natural language processing, and computer vision, giving him rare versatility.

Ha’s publication portfolio is especially noteworthy. He is an author or co-author on over 100 papers at top-tier international AI conferences (ICLR, CVPR, NeurIPS, ICML, EMNLP, among others), many highly cited—his h-index is over 40, with some of his papers cited thousands of times. Significant works include advances in continual learning (e.g., “Rainbow Memory: Continual Learning With a Memory of Diverse Samples”), generative adversarial networks (“StarGAN: Unified Generative Adversarial Networks for Multi-Domain Image-to-Image Translation”), and practical applications of large language models (LLMs), including the landmark “What Changes Can Large-scale Language Models Bring? Intensive Study on HyperCLOVA” which details the technical design and benchmarks for Naver’s sovereign LLM.

Ha’s academic contributions extended beyond research. His leadership in collaborative research initiatives—most notably the SNU-Naver Super-large AI Center—illustrates a commitment to linking Korean academia and industry on a scale typically dominated by Western or Chinese technology ecosystems. Recognition of his work features in international benchmarking: under his leadership, Naver’s AI research impact reportedly reached sixth globally in metrics for conference publications and reputation.

Leadership Roles at Naver

Ha Jung-woo’s move into industry was marked by a series of escalating leadership positions. He joined Naver Labs (the research and development arm of Korea’s leading internet conglomerate) in 2015, just as the company began pivoting aggressively into AI. From 2017 to 2019, he led the CLOVA AI research team—a key node in Naver’s strategy to build AI services in language, search, and recognition, laying the groundwork for Korean-centric conversational AI.

By October 2020, Ha had been appointed head of Naver AI Lab, overseeing both mid- and long-term advanced AI research. Notably, he led cross-institutional collaborations between Naver and major Korean research universities like KAIST and SNU, fostering integrated AI research pipelines that positioned Korea to compete with U.S. and China on sovereign LLMs and core infrastructure.

As Executive Vice President of Naver Cloud and Head of the company’s AI Innovation Center, Ha’s portfolio grew to encompass both research leadership and strategic planning. Under his stewardship, Naver launched Korea’s first hyperscale AI model, and he orchestrated R&D teams developing technologies for natural language processing, multimodal perception, and LLM safety. The outcome was not just academic—Ha drove the successful commercialization of AI services, such as the rollout of Naver’s HyperCLOVA X across cloud, search, enterprise, and public sector channels.

Development of HyperCLOVA X

HyperCLOVA X, launched under Ha’s leadership, represents the most ambitious Korean-led large language model to date. Designed to rival models like GPT-4 and Google’s Gemini, HyperCLOVA X is a family of LLMs specifically tailored for the Korean language and cultural context, but also competitive in English, mathematics, and coding.

Key Features

• Scale and Data: HyperCLOVA X was trained on a balanced corpus triple—Korean, English, and code—using over 204 billion parameters, exceeding GPT-3’s 175B and dwarfing earlier Korean models. It uses 6,500 times more Korean data than GPT-4, aiming for cultural alignment.
• Cultural and Ethical Context: The model is imbued with knowledge specific to Korean history, laws, values, and societal issues, filling crucial gaps left by American or Chinese LLMs (where, for instance, concepts like the “East Sea” or certain sensitive social contexts might be misunderstood or misrepresented).
• Performance Benchmarks: HyperCLOVA X outperforms all open-source and many closed-source models (including GPT-4) on Korean-centric benchmarks, while matching the best global alternatives in English-centric tests. Its benchmarks include custom datasets (e.g., KoBigBench, KMMLU, HAE-RAE) and global ones (MMLU, BigBench-Hard).
• Instruction Following and Safety: Methods such as supervised fine-tuning and reinforcement learning from human feedback were employed, with an explicit safety framework and red-teaming protocols to minimize harmful, biased, or culturally inappropriate outputs.
• Applications: Beyond research, HyperCLOVA X powers Naver's enterprise services (cloud, chatbots, translation, multimedia search), Seoul city’s digital public services, and supports educational and business AI across the country.

Notably, HyperCLOVA X is not just a commercial or technological achievement; it is explicitly positioned as the referential blueprint for sovereign AI—its technical reports, benchmarks, and safety frameworks are shared openly to help other regions or countries develop their own LLMs, putting Korea at the forefront of global AI ecosystem-building.

Sovereign AI Concept and National Blueprint

Dr. Ha Jung-woo’s most influential policy and conceptual contribution is the formulation of “sovereign AI”—an idea that has defined South Korea’s current national strategy. Unlike generic LLMs developed in the U.S. or China, sovereign AI for Ha is not merely about model ownership but about national autonomy over the entire AI value chain: from GPUs and semiconductors, to language data, models, infrastructure, and ethical governance.

Philosophical Underpinnings and Policy Objectives

• Technological Independence: Sovereign AI addresses Korea’s overreliance on foreign infrastructure. Ha advocates for domestic production or strategic acquisition of computing hardware (notably GPUs), government-subsidized data centers, and public investment in foundational AI R&D.
• Cultural and Legal Alignment: A sovereign AI must internalize a nation’s culture, laws, and values, guaranteeing that public sector and sensitive applications (defense, healthcare, welfare) are deployed with full legal and social context.
• Public-Private Virtuous Cycle: Echoing Ha’s repeated mantra, sovereign AI should foster a cyclical strategy: the state underwrites core infrastructure and foundational R&D, while corporations productize and commercialize, and then share achievements and datasets to the broader public sector.
• Open-Source and Ecosystem Building: Unlike proprietary models, Ha insists that sovereign AI built with public funds should be offered as open-source, ensuring that the innovation benefits are dispersed across startups, educational institutions, and the social economy.

The sovereign AI concept is not insular; Ha frames it as a “third way” in the global tech rivalry—where Korea serves as a strategic partner for middle-power countries in Southeast Asia, the Middle East, and beyond, providing co-development, technology transfer, and pledges of data and digital sovereignty that the U.S. and China might not offer.

AI Ethics, Social Responsibility, and Civic Engagement

Ethics and social responsibility form a foundational layer in Ha Jung-woo’s AI vision. Cognizant of the risks posed by powerful LLMs, Ha has advocated for responsible AI that balances innovation with the needs, rights, and values of society.

Structural Initiatives

• AI Ethics Principles: The HyperCLOVA X project incorporated systematic “red teaming,” strict data curation for privacy and fairness, and published explicit AI Ethics Principles—covering categories such as non-discrimination, privacy, avoidance of hate or misinformation, and the prohibition of harmful content.
• Public Engagement: As co-representative of the National Coalition for the Realization of a Right Science and Technology Society and as a frequent public lecturer, Ha emphasized “ecosystem design” and the need for a governance model combining state oversight, private dynamism, and civic values. He chaired or participated in over 800 public lectures and forums in 2024 alone, demystifying AI policy for the public and industry.
• AI Future Forum and Civic Dialogue: As co-chair of the AI Future Forum (backed by the Science and Technology Society Realization Alliance), Ha organized cross-party and multi-stakeholder dialogues to surface societal risks and to inject real-world concerns—on labor impact, climate change, education, and startup innovation—into technological policy debate.

The Governance Challenge

Ha has been forthright about Korea’s systemic hurdles: a legacy of bureaucratic inertia, lack of agile startup support, and rigid regulatory frameworks hindering ecosystem innovation. He argues that without collaborative governance and a focus on protecting social interests, AI progress risks becoming a “first-in-technology, last-in-accountability” phenomenon.

Advisory and Committee Roles in Government

Before his presidential appointment, Ha Jung-woo accumulated a wide array of advisory roles, spanning conservative and progressive administrations, and government ministries:

• Presidential Digital Platform Government Committee: As chair of the AI-Data Subcommittee and leader of the Task Force for Super-Scale Public AI, Ha advised on projects integrating public AI services, open data ingests, and responsible scaling of public AI infrastructure under the Yoon Suk-yeol government.
• Financial and Strategic Technology Councils: Ha served as IT Subcommittee advisor for the Financial Supervisory Service’s Advisory Committee and as a member of the National Strategic Technology Committee (Ministry of Science and ICT) and the Strategic Planning and Investment Council (Ministry of Trade, Industry and Energy), giving him broad insight into the intersections of AI with finance, industry, and technology management.
• Public Forums and Party Outreach: In the volatile political climate after Yoon’s impeachment, Ha actively engaged with the then-Democratic Party (appearing with party leader Lee Jae-myung on YouTube) and later as a key panelist at “The Age of AI: Discovering a New Path for South Korea,” shaping the party’s campaign promises on AI.

This history of bipartisan and cross-ministerial engagement uniquely qualified Ha for the “control tower” position created under President Lee Jae-myung.

AI Future Forum and Broader Civic Engagement

The AI Future Forum, which Ha co-chaired for four years, emerged as a crucial civic platform in Korean AI governance. Rooted in the Science and Technology Society Realization Alliance, the forum achieved:

• Extensive Outreach: Over 800 lectures and workshops in 2024, directly engaging with scientists, policymakers, entrepreneurs, and ordinary citizens.
• Cross-Sector Policy Input: The forum convened engineers, lawyers, ethicists, educators, and businesspeople to build consensus on ethical AI, data governance, job displacement, and privacy. Outcomes were fed into government white papers and legislative advisories.
• Talent Cultivation: Ha spearheaded bootcamps and special programs—such as his tenure as principal of Gwangju AI High School—focused on developing a new generation of Korean AI practitioners, particularly targeting underrepresented youth outside Seoul.

This engagement ensured the AI strategy was deeply tied to issues of public trust, talent pipeline, and social cohesion.

Table: Key Milestones in Ha Jung-woo’s Career and AI Policy Contributions

Each of these milestones is marked by a drive for integration: of research and application, public and private funding, technical and social progress, and ultimately, sovereignty within the global context.

Appointment as Senior Presidential Secretary for AI Future Planning

Ha Jung-woo’s appointment on June 15, 2025, as the country’s inaugural “Senior Presidential Secretary for AI Future Planning” is a watershed moment. The role, newly created by President Lee Jae-myung, is conceived as the central “control tower” for making South Korea a top-three global AI powerhouse—a campaign pledge now operationalized through specific investment and talent commitments.

Mandate and Authority

• Investment Oversight: The Secretary is charged with managing the rollout of a 100 trillion won (approx. $73–$75 billion USD) public/private AI investment program, including funding for national data centers, hardware procurement (notably GPUs), and talent programs for training 100,000+ AI professionals.
• Policy Leadership: Ha heads the cross-ministerial AI planning committee, coordinates the Science/Tech Secretary, population policy advisors, and climate/energy policy-makers—integrating AI with industrial, demographic, and climate strategies.
• Public-Private Bridge: Drawing on his private sector and civic experience, Ha’s mandate explicitly centers on translating “field knowledge” from industry and academia into national policy—correcting prior bureaucracy- or vendor-centric strategies.
• International Diplomacy: Ha represents Korea internationally, advocating for alliances among middle-power countries to mitigate U.S./China AI hegemony and to promote Korean AI as a platform for regional partners in Asia and beyond.

The appointment drew wide support and expectation from industry, but also raised debate about the risk of market concentration and the need for transparent, market-oriented public-private models.

Role in National AI Strategy—Lee Jae-myung Administration

Under the Lee administration, AI future planning has become the lodestone of government economic policy and is deeply interwoven with Korea’s aspiration for economic, demographic, and strategic renewal.

Strategic National Objectives

• Economic Rejuvenation: AI is central to restoring Korea’s potential growth rate (targeting 3% GDP growth) as the country faces long-term demographic decline and investment slowdown. The Lee administration frames AI transformation as akin to wartime mobilization or an “emergency national effort”.
• Flagship Projects: The policy blueprint centers on four “AI flagship projects”: (1) scaling national AI computing infrastructure (expanding GPU capacity 15-fold by 2030 and building 2 exaflops+ computing centers), (2) large-scale private and public AI investment (targeting 100 trillion won), (3) AI+X industry transformation and public sector adoption (aiming for 95% public institution AI use), and (4) establishing global standards for AI safety/ethics.
• World Best LLM Project: Under Ha’s direction, Korea is racing to develop an open-source “World Best” large language model, with a national open competition, hardware grants, and data pipe-lining for sovereign public use—explicitly countering Big Tech’s proprietary models.
• Global Engagement: Korea, under Ha’s external advocacy, seeks to anchor a coalition of technology-sharing, AI-sovereign middle powers, extending its model into diplomatic and international cooperation forums such as the G7 and APEC.

Organizational Leadership

Ha chairs or co-chairs multiple cross-ministerial committees, including:

• AI-Data Subcommittee: Steers public data policy, infrastructure planning, and privacy standards.
• Super-Scale Public AI Task Force: Directs coordination of public AI services, open-source LLM distribution, and interfaces with cloud players like Naver and SK Group.
• National AI Planning Committee: The Committee, now under an independent secretariat, is Korea’s strategy clearinghouse—rolling out implementation plans for all national AI investments.

Chairing AI-Data Subcommittee and Super-Scale Public AI Task Force

Dr. Ha’s experience as chair of the AI-Data Subcommittee and head of the Task Force for Super-Scale Public AI has been critical to South Korea’s rise as an AI-first digital society.

Subcommittee Achievements

• Open Data Governance: Advanced integration of public datasets (urban, educational, medical) into AI training corpora, ensuring both privacy-preserving open data and fair access for startups and academia.
• Super-Scale Infrastructure: Designed the digital backbone (storage, compute, networking) for AI services powering Seoul's city governance, national smart health, and education platforms.
• Cloud and GPU Strategy: Directly oversaw policy recommending the government bulk-purchase thousands of top-of-the-line GPUs and subsidize infrastructure for designated national AI champions (Naver, SK, others), in line with the “full-stack” sovereignty doctrine.

Policy Challenges

Despite ambition, not all initiatives were smooth. The “National AI Computing Center” project—modeled on U.S. “Stargate” data center plans—faced two failed bids due to misalignment between public incentives and commercial realities, revealing the need for deeper market consultation and adaptive public policy.

Impact, Critique, and the Road Ahead

Dr. Ha's stewardship is already reshaping Korea’s global technology ambitions and internal innovation systems, but not without tension and debate.

Impact

• Global Recognition: Korea now ranks third worldwide in the number of proprietary generative AI models, trailing only the U.S. and China, and is taken seriously as a sovereign AI pioneer in international fora.
• Economic Ecosystem: The AI transformation agenda is triggering new investment, university-industry partnerships, public sector reform, and a startup ecosystem focused on both vertical AI and foundational technologies.

Critique and Debate

• Market and Innovation Risk: Some Korean startups and AI entrepreneurs worry that the strategy inevitably favors chaebols (conglomerates), and that the drive for “big is beautiful” LLM investments crowds out vertical, market-driven, or open innovation. There is increasing demand for transparent governance and equal access to public investment benefits.
• Talent and Globalization: Korea still faces global competition for AI talent (many top OpenAI and Google AI researchers are Korean expatriates), and there are questions about whether compensation and R&D freedom can match U.S. or Chinese tech ecosystems.
• Operational Bottlenecks: GPU and hardware shortages, data privacy laws, and bureaucratic inertia remain obstacles, as do persisting failures in large public procurement schemes (National AI Computing Center) without adaptive, startup-friendly frameworks.

Conclusion

Dr. Ha Jung-woo encapsulates the new face of Korean techno-civil strategy: a researcher-bureaucrat who moves fluidly between laboratory leadership, industry-scale development, and the public sphere. Under his leadership, sovereign AI is not only a technical ambition but a policy and ethical vision—one that fuses innovation, social responsibility, and global diplomacy into a single blueprint.

The true test of this experiment will be in the years ahead: Can Korea’s sovereign AI platform sustain public and private innovation, balance market and public-good dynamics, and secure a seat at the global table as both a technological and ethical leader? With Ha at the helm, Korea is uniquely positioned. If successful, his approach may serve as a template for other middle-power nations navigating the AI revolution between the gravitational pulls of Washington and Beijing.

References to Milestones, Initiatives, and Impact

The report integrates insights and direct data from diverse English- and Korean-language sources, including top business and policy newspapers, official government statements, technical reports from Naver and the Ministry of Science and ICT, scholarly databases, AI technical benchmarks and performance tables, interviews, and trade press analyses. Notable references include CHOSUNBIZ, The Korea Times, BusinessKorea, Korea IT Times, NamuWiki, Blockmedia, The Diplomat, Korea JoongAng Daily, KED Global, and the MSIT/AI National Committee announcements, as well as open technical repositories and Google Scholar indices.

Table: Summary of Key Milestones in Ha Jung-woo’s Career and AI Policy Contributions

By charting this journey, Dr. Ha Jung-woo's impact emerges as both visionary and operational, melding the best of Korea’s academic, corporate, and public policy domains to pioneer a new model for sovereign, ethical, and globally relevant AI leadership.

References

1. Jung-Woo Ha - Google Scholar. https://scholar.google.com/citations?user=eGj3ay4AAAAJ&hl=en
2. Ha Jung -woo (1977) - NamuWiki. https://en.namu.wiki/w/%ED%95%98%EC%A0%95%EC%9A%B0%281977%29
3. Jung-Woo Ha - OpenReview. https://openreview.net/profile?id=~Jung-Woo_Ha1
4. HyperCLOVA X Technical Report - arXiv.org. https://arxiv.org/pdf/2404.01954
5. President Lee Appoints Naver’s Ha Jung-woo as Inaugural AI Future .... https://www.businesskorea.co.kr/news/articleView.html?idxno=244634
6. Naver eyes global market with new LLM HyperCLOVA X. https://www.kedglobal.com/artificial-intelligence/newsView/ked202308240018
7. HyperCLOVA X . https://clova.ai/en/hyperclova
8. South Korea’s Sovereign AI Gambit: A High-Stakes Experiment in Autonomy. https://thediplomat.com/2025/07/south-koreas-sovereign-ai-gambit-a-high-stakes-experiment-in-autonomy/
9. South Korea Unveils $735 Billion Plan for Sovereign AI Leadership. https://editorialge.com/south-korea-sovereign-ai-735-billion-investment-strategy/
10. Korea’s New Path in AI - Can Ha Jung-Woo Bridge Innovation and Social .... https://www.koreaittimes.com/news/articleView.html?idxno=142436
11. President Lee appoints AI expert Ha Jung-woo to lead national AI .... https://biz.chosun.com/en/en-it/2025/06/15/AXUBDKEGZZC55OJRFMFO5SYOEU/
12. Ha Jung-woo proposes 'sovereign AI' blueprint to boost South Korea's AI .... https://biz.chosun.com/en/en-it/2025/06/16/P3HGWTIOUVAGROFOVHRMDYXG4I/
13. Tech expert from Naver named inaugural senior secretary for AI. https://www.koreatimes.co.kr/southkorea/politics/20250615/tech-expert-from-naver-named-inaugural-senior-secretary-for-ai
14. Korea to Invest 100 Trillion Won in Projects to Lead Super Innovation. https://www.businesskorea.co.kr/news/articleView.html?idxno=250086
15. Press Releases: Korea.net : The official website of the Republic of Korea. https://www.korea.net/Government/Briefing-Room/Press-Releases/view?articleId=7606&type=O
16. South Korea's AI Strategy Targets World Leadership with Major .... https://www.businesskorea.co.kr/news/articleView.html?idxno=235833
17. Press Releases - 과학기술정보통신부 >. https://www.msit.go.kr/eng/bbs/view.do?sCode=eng&mId=4&mPid=2&bbsSeqNo=42&nttSeqNo=1040&searchOpt=ALL
18. On DeepSeek's shock waves: A conversation with Ha Jung-woo, Korea's .... https://koreajoongangdaily.joins.com/news/2025-02-12/opinion/columns/On-DeepSeeks-shock-waves-A-conversation-with-Ha-Jungwoo-Koreas-leading-AI-expert/2240506
19. Naver's Ha Jung-woo appointed as South Korea's first AI Future Planning .... https://biz.chosun.com/en/en-policy/2025/06/15/QAUA7MCDIZCYDCRORFLPRXXNGE/