Institute for Information & Communications Technology Planning & Evaluation (IITP)

Mandate and Institutional Role

The Institute for Information and Communications Technology Planning and Evaluation (IITP) is the Korean government agency responsible for planning, managing, and evaluating information and communications technology (ICT) research and development programmes. Operating under the supervision of the Ministry of Science and ICT (MSIT), IITP serves as the operational engine for Korea's AI development strategy and is the primary executing agency for the technology-focused components of K-Moonshot. Its mandate places it at the critical juncture where government policy directives are translated into funded research programmes, evaluated for technical merit, and monitored for progress toward measurable outcomes.

While the National Research Foundation (NRF) funds basic scientific research and KIAT manages industrial technology commercialisation, IITP occupies the decisive middle ground: applied ICT R&D that transforms fundamental research into functional technologies, platforms, and systems. In the K-Moonshot context, this means IITP manages the programmes developing Korea's sovereign AI foundation models, AI computing infrastructure, next-generation communications systems, AI-powered software platforms, and the evaluation frameworks that determine which AI projects receive continued government funding.

IITP was established in 2013 through the reorganisation of functions previously distributed across the Institute for Information Technology Advancement (IITA) and portions of the National IT Industry Promotion Agency (NIPA). The consolidation aimed to create a more streamlined R&D management system for Korea's rapidly expanding ICT sector, which generates approximately 15% of Korean GDP and forms the technological backbone of the K-Moonshot architecture. The agency's professional staff includes programme managers with doctoral-level expertise across ICT domains, technology evaluators, policy analysts, and specialists in intellectual property management and technology transfer.

ICT R&D Programme Management Lifecycle

IITP manages an annual ICT R&D portfolio exceeding 3 trillion won, distributed across more than 500 active projects. The agency's management model follows a systematic lifecycle that gives it influence at every stage of the R&D process:

• Technology Roadmapping: IITP conducts annual technology roadmapping exercises that identify strategic priorities, capability gaps, and research opportunities across the ICT sector. These roadmaps directly inform MSIT's R&D budget allocation decisions and set the research agenda for competitive calls for proposals. Under K-Moonshot, the roadmapping process has been restructured to explicitly map technology requirements against the 12 national mission objectives.
• Programme Design: Based on technology roadmaps and MSIT policy directives, IITP designs specific R&D programmes with defined objectives, timelines, budgets, and performance metrics. K-Moonshot has significantly influenced programme design, with new programmes explicitly aligned to mission requirements and structured to demand inter-institutional collaboration.
• Project Selection and Evaluation: IITP manages competitive calls for proposals, organises expert review panels drawn from academia and industry, and selects funded projects. The selection process emphasises the feasibility of proposed technical approaches, the qualifications of research teams, and the potential for technology transfer and commercialisation. This is where IITP exercises its most consequential gatekeeping function: determining which AI research proposals merit government investment.
• Monitoring and Milestone Tracking: Active projects undergo regular progress reviews against defined milestones and deliverables. IITP employs both quantitative metrics (publications, patents, working prototypes, benchmark performance) and qualitative expert assessments to evaluate whether projects are progressing satisfactorily.
• Performance Evaluation: Completed projects are evaluated for scientific output, technology transfer, commercialisation outcomes, and human capital development. Evaluation results feed directly back into future programme design and funding decisions, creating an institutional learning loop.

Sovereign Foundation Model Consortia

IITP's highest-profile K-Moonshot responsibility is the management of five government-funded consortia competing to develop Korea's sovereign AI foundation models, backed by a total investment of approximately 500 billion won. These consortia represent the operational core of Mission 7: Physical AI Models and constitute one of the most ambitious government-directed AI development programmes outside the United States and China.

Each consortium includes university research groups, specialised AI companies, and in some cases government-funded research institutes, creating multi-institutional teams that combine corporate engineering resources with academic research depth. IITP's role encompasses setting technical milestones, coordinating GPU access through the National AI Computing Center, organising standardised benchmark evaluations, ensuring alignment with Korea's AI ethics framework, and managing the complex intellectual property arrangements between consortium members with competing commercial interests.

The five-consortium model is a deliberate design choice. Rather than selecting a single national champion, IITP has structured the programme to foster competition among Korea's major technology companies while ensuring that each consortium tackles a distinct application domain. This approach distributes risk across multiple technical approaches but also raises the question of whether Korea's AI research capacity is sufficient to sustain five parallel world-class foundation model efforts simultaneously. IITP's ongoing evaluation of consortium performance will ultimately determine whether the competitive model is consolidated or maintained.

AI Computing Infrastructure Coordination

IITP is the implementing agency for MSIT's GPU deployment programme, managing the technical dimensions of Korea's AI computing infrastructure build-out. This role makes IITP responsible for one of the most tangible elements of K-Moonshot: ensuring that Korean AI researchers, startups, and companies have access to the computational resources necessary to train and deploy competitive AI models.

• National AI Computing Center: IITP manages the allocation of GPU resources from the national pool, prioritising access for sovereign model development, academic research, and startup computing needs. The initial deployment of 4,000 GPUs from the 10,000-unit pool began in March 2026, with demand management and access scheduling handled by IITP's infrastructure team.
• Sovereign Cloud Partnership Management: IITP coordinates the technical and contractual dimensions of sovereign cloud partnerships with Naver Cloud, Kakao, NHN Cloud, and KT Cloud, ensuring that government-backed AI computing infrastructure meets the performance, security, and data sovereignty requirements specified by MSIT.
• Korean-Language AI Data Infrastructure: IITP manages programmes to build Korean-language training datasets, data labelling platforms, and data governance frameworks. Sovereign AI model development requires large-scale, high-quality Korean-language corpora that reflect Korean cultural context, legal terminology, and domain-specific knowledge. IITP coordinates the creation of these datasets across government, academic, and private-sector data sources.
• Benchmark and Evaluation Frameworks: IITP has developed Korean-language AI benchmark suites that evaluate model performance on tasks specific to the Korean language and cultural context, addressing the limitations of English-centric international benchmarks. These benchmarks provide the standardised evaluation criteria against which sovereign model consortia are measured.

AI Chip Development Support

IITP manages R&D programmes directly relevant to Mission 11: Ultra-High-Performance AI Accelerators. The agency funds research in neural processing unit (NPU) architectures, AI-specific compiler development, chip-model co-optimisation, and the development of Korean AI semiconductor design tools. Key funded entities include Rebellions, FuriosaAI, Sapeon (an SK Telecom subsidiary), and university research groups at KAIST and POSTECH developing next-generation AI hardware architectures.

The AI chip programme reflects Korea's strategic imperative to reduce dependence on foreign AI accelerator suppliers, particularly NVIDIA, whose GPUs currently dominate AI training and inference workloads globally. While Korea possesses world-leading semiconductor manufacturing capability through Samsung Foundry and SK Hynix's HBM leadership, the design of competitive AI accelerator chips requires distinct competencies in architecture, compiler technology, and software ecosystem development. IITP's R&D programmes aim to build these capabilities domestically.

Beyond AI: Full ICT R&D Portfolio

While AI dominates IITP's current strategic agenda, the agency manages R&D programmes across the full ICT technology stack that provide essential infrastructure for K-Moonshot's digital ambitions:

• 6G Communications: Korea aims to commercialise 6G technology by 2028-2030. IITP manages pre-standardisation research programmes involving Samsung, LG, SK Telecom, and KT, addressing terahertz spectrum utilisation, AI-native network architectures, and satellite-terrestrial integration for non-terrestrial networks.
• Software Engineering and Cloud: IITP supports Korean software industry development through programmes funding cloud-native architectures, DevOps platforms, enterprise software products, and open-source ecosystem participation.
• Cybersecurity: As Korea's AI infrastructure expands, IITP manages R&D in AI-powered cybersecurity, zero-trust architectures, and secure AI model deployment, all critical for protecting the national AI computing infrastructure from state-sponsored and criminal threats.
• Data Platforms: Programmes supporting data marketplace technology, privacy-preserving analytics tools, and federated learning infrastructure that enable AI training on sensitive data while complying with Korea's Personal Information Protection Act (PIPA).

Technology Readiness and Commercialisation Bridge

IITP's distinctive position in Korea's R&D funding architecture lies in its focus on technology readiness levels 3 through 7, bridging the gap between basic laboratory discovery funded by the NRF and the industry-led commercialisation programmes managed by KIAT. This bridging function is particularly important for K-Moonshot, which demands not merely scientific breakthroughs but functional, deployable technologies ready for industrial adoption.

IITP enforces this orientation through several mechanisms. Most IITP-funded projects require industry partners who contribute matching funds and commit to commercialising research outcomes. Projects must include milestones for prototype development, system demonstration, and pilot deployment. The agency works with university and GRI technology transfer offices to facilitate licensing of IITP-funded technologies to Korean companies, including K-Moonshot Corporate Partnership members. Increasingly, IITP requires that funded projects contribute to open-source repositories and participate in international standards development.

Coordination Across the Funding Triad

IITP operates within a three-agency funding model alongside NRF and KIAT, with overlapping TRL ranges that create both productive tension and coordination challenges. Under K-Moonshot, MSIT has established inter-agency coordination committees with IITP playing a central connecting role, ensuring that ICT-related research progresses smoothly from basic science through applied development to industrial application. The effectiveness of this coordination will significantly determine whether K-Moonshot's AI investments produce world-class outcomes or encounter the translation gaps that have plagued previous Korean R&D programmes.

Strategic Assessment

IITP is arguably the most operationally consequential agency in K-Moonshot's AI and ICT domain. While MSIT sets strategic direction and NRF funds upstream science, IITP manages the applied R&D programmes that will determine whether Korea can develop competitive sovereign AI models, build sufficient computing infrastructure, produce domestic AI chips, and train the next generation of AI engineers.

The sovereign foundation model programme is IITP's highest-stakes initiative. Whether five competing consortia produce better outcomes than a more concentrated investment remains an open question that IITP's programme management decisions will help answer. The agency's evaluation capability, its ability to distinguish genuine technical progress from superficial milestone compliance in a domain where capabilities advance rapidly and benchmarks become obsolete quickly, will determine whether K-Moonshot's AI investments produce world-class capabilities or merely incremental improvements over existing Korean AI platforms.

IITP's programme management capacity itself is a potential bottleneck. The substantial increase in AI and ICT R&D funding under K-Moonshot requires corresponding increases in programme management capability. Recruiting programme managers with the expertise to evaluate cutting-edge AI research, negotiate complex consortium agreements, and manage multi-billion-won technology programmes is challenging when the private sector competes aggressively for the same talent. IITP's ability to attract and retain qualified programme managers may ultimately prove as important to K-Moonshot's success as the technical research itself.

For analysis of Korea's sovereign AI strategy, see AI Sovereignty. For the semiconductor and AI chip landscape, see the Semiconductor Sector and HBM Dominance analyses. For IITP's relationship with the broader K-Moonshot budget, see Korea AI Budget 2026.