South Korea - National AI Strategy

Overview

The National Strategy for Artificial Intelligence (established 17 December 2019) is South Korea’s pan‑government strategic roadmap setting the vision 'Toward AI World Leader beyond IT' for 2030. The Strategy organizes 9 strategies and 100 action tasks across three pillars—AI ecosystem, AI utilization and people‑centered AI—and assigns responsibilities across ministries and implementing agencies. The strategy emphasizes (i) scaling AI infrastructure and data platforms, (ii) accelerating AI adoption across industry and government, and (iii) ensuring human‑centric, trustworthy AI. The primary official announcement and summary are published by the Ministry of Science and ICT; the official policy page is available at MSIT – National Strategy for Artificial Intelligence and a downloadable copy of the full strategy text is available (archived) at National Strategy for Artificial Intelligence (PDF).

Definitions

The Strategy adopts operational definitions used across government planning documents: 'AI' is defined broadly to include machine learning, deep learning and data‑driven decision‑support systems that perform cognitive tasks such as perception, reasoning and language understanding. 'Data' covers public and private datasets used for AI training and service delivery; the Strategy uses 'data platform' to mean government or sectoral repositories, access APIs and curated datasets for research and innovation. 'People‑centered AI' refers to AI development and deployment that respect human dignity, privacy and fairness and that prioritize societal benefit and safety. The document distinguishes between strategic 'action tasks' (non‑binding, assigned to ministries) and regulatory instruments (laws or ministerial rules), making clear that the Strategy itself is policy guidance rather than primary legislation.

Governance and Institutional Framework

Governance is multi‑layered: the Presidency and the Presidential Committee on the Fourth Industrial Revolution (PCFIR) provide high‑level direction and coordination, MSIT acts as the lead ministry for AI policy, and implementation is distributed among line ministries and specialized agencies such as the National Information Society Agency (NIA) and the South Korea Internet & Security Agency (KISA). The Strategy mandates pan‑government coordination, assigns lead ministries for each of the 100 tasks, and requires ministries to incorporate tasks into annual work plans. It also calls for the formation of inter‑ministerial working groups, sectoral 'AI‑X' project teams, and public‑private councils. For governance and public materials see MSIT’s policy page at MSIT policy page and the Presidential Committee portal at Presidential Committee on the Fourth Industrial Revolution (English). The Strategy envisages periodic strategic meetings chaired by the President or senior ministers to review performance and to mobilize budgetary resources.

Key Focus Areas

The Strategy concentrates on nine sub‑strategies grouped into three core areas: (A) Establishing a global‑leading AI ecosystem—(1) full opening and linking of public data, (2) building AI learning datasets and data platforms, (3) investing in AI infrastructure (cloud, HPC, and edge), (4) securing semiconductor and AI chip competitiveness; (B) The country that best uses AI—(5) accelerating AI adoption across manufacturing, health, transport, energy and services (AI+X projects), (6) transforming public services and implementing a digital government with AI‑driven services; and (C) Realizing people‑centered AI—(7) promoting AI ethics and trustworthy AI standards, (8) human capital and lifelong learning programs for AI skills, and (9) social safety nets and labor market policies for AI‑driven transitions. Specific tasks include establishing national data platforms, expanding publicly funded AI R&D, launching regional AI demonstration projects, creating graduate AI schools and vocational retraining, and promoting voluntary ethics codes. The Strategy balances economic competitiveness (e.g., next‑gen chips and industry support) with non‑economic aims (privacy, fairness and inclusion). Implementation relies heavily on public investment, cross‑sector partnerships, and standards/ethics development to enable adoption while addressing social concerns.

Implementation Framework

Implementation is framed as a distributed, ministerial responsibility model. Each of the 100 tasks is allocated to a lead ministry or agency with timelines and performance indicators to be integrated into ministry annual plans. The Strategy calls for targeted public investments (R&D funding, infrastructure budgets), creation of national datasets (open by default), and pilot programs (AI sandboxes, AI Grand Challenge) to accelerate commercialization. Regulatory innovation is promoted—where appropriate, the Strategy encourages regulatory sandboxes and regulatory reviews to remove outdated rules that block AI innovation ('approve first, regulate later' in pilot contexts). MSIT is the coordinating point; for finance and budgets the Ministry of Economy and Finance is engaged. Agencies such as NIA and KISA are tasked with technical operations (data platform management, cybersecurity). The Strategy also outlines mechanisms for ethical oversight, voluntary certification, and standards development to support interoperability and trust.

Monitoring and Evaluation

The Strategy specifies periodic monitoring mechanisms: inter‑ministerial reviews, performance reporting into presidential strategic meetings, and public disclosure of progress indicators. It envisages annual status reports and the publication of major milestones. Ministries must report on the integration of the Strategy’s action tasks into their work plans, and designated agencies will collect metrics (e.g., dataset volumes, AI R&D spending, number of trained personnel, pilot outcomes). The Strategy emphasizes transparency of performance data to stimulate public‑private accountability and to guide budgetary adjustments. While the Strategy lacks statutory enforcement mechanisms, it relies on executive oversight and budgetary levers to maintain momentum.

Penalties, Liability, and Appeals

As a strategic policy document (not primary legislation), the Strategy does not itself prescribe criminal penalties or administrative fines. Instead, it references that legal compliance (privacy, safety, consumer protection) remains subject to existing laws such as the Personal Information Protection Act (PIPA) and sectoral safety and consumer laws. Liability and redress for harm caused by AI are to be handled under existing civil and administrative frameworks; the Strategy calls for follow‑on work to clarify liability questions as AI systems become more complex. For enforcement and penalties, implementing agencies are expected to coordinate with regulators that hold statutory authority (privacy commission, financial regulators, health regulators) to ensure applicable sanctions under existing law are applied where violations occur.

Relationship to Other Instruments

The Strategy is explicitly designed to align with and complement existing Korean laws and policies: the Personal Information Protection Act, the Act on Promotion of Information and Communications Network Utilization and Information Protection, credit and financial laws for data use, and sectoral safety and medical device regulation. It also builds on prior government plans (AI R&D Strategy 2018, Data Industry Vitalization Strategy 2018, System Semiconductor Strategy) and anticipates the enactment or revision of data‑focused legislation to support broader data use. The Strategy functions as an umbrella policy that guides ministry‑level rules, programs, and standards rather than replacing them.

International Alignment

South Korea situates its national approach within international AI policy dialogues—contributing to OECD AI policy work and promoting alignment with global AI ethics and standards. The Strategy calls for international cooperation in R&D, standard setting, and trusted partnerships (particularly around semiconductors, cloud infrastructure and AI safety). South Korea’s approach emphasizes interoperability with international norms while leveraging national strengths (semiconductor manufacturing, ICT infrastructure). For international engagement, the Strategy references bilateral and multilateral cooperation channels and cross‑border data policy coordination. See MSIT and Presidential Committee materials for South Korea’s international AI cooperation agendas (MSIT policy page, PCFIR).

Implementation Timeline

Milestone	Date	Notes
Presidential Initiative announced (DEVIEW/Oct 2019)	2019-10-28	Presidential initiative that framed the national strategy.
National Strategy adopted (Cabinet)	2019-12-17	Government‑wide adoption and assignment of 100 action tasks.
Ministries integrate tasks into 2020 annual plans	2020-01 to 2020-12	Annual planning and first year of implementation.
Mid‑term review / ethics standards publication	2020-2021	Ethics guidance and trusted AI strategy follow‑ons.

Compliance Checklist

Requirement	Who	Checklist
Integrate assigned action tasks into ministry annual plan	Line ministries	Task assignment, budget allocation, KPIs set
Publish/open public datasets	Public agencies	Data cataloguing, anonymization, API access
Adopt AI ethics and conduct impact assessments	AI developers and public agencies	Ethics code adoption, transparency disclosures, logging
Participate in pilot sandboxes and certification	Industry partners	Apply to sandbox, share results, pursue voluntary certification

Sources and References

Source	Type
Ministry of Science and ICT — National Strategy for Artificial Intelligence (policy page)	Primary Source
National Strategy for Artificial Intelligence (full text, PDF, archived)	Primary Source