🌱 Welcome to CUBIG – shaping the future of safe and intelligent public data

As governments worldwide shift toward AI-driven administration, the demand for public data that is both safe and immediately usable for machine learning continues to grow.

CUBIG is excited to introduce how DTS, our secure synthetic data infrastructure, is becoming a critical enabler for this transformation. With public institutions now required to open higher-value, AI-friendly datasets, synthetic data is emerging as the most practical path forward.

DTS supports this movement by ensuring that sensitive information never leaves the organization while still generating data suitable for analytics, policy modeling, and AI services.

🔐 A Policy Shift Toward Quality and AI-Readiness

The Ministry of the Interior and Safety recently surpassed 102,000 datasets on the Public Data Portal and announced a transition from large-scale disclosure toward high-value, AI-ready data.

New evaluation criteria introduced in 2025 include efforts to provide AI-friendly datasets and performance in pseudonymized or synthetic data opening. This effectively positions synthetic data as a requirement rather than an option.

However, many of the most valuable datasets—such as resident registration, healthcare, welfare, and civil complaints—cannot be opened due to privacy risks. In response, regulatory bodies including the Personal Information Protection Commission have issued guidelines endorsing synthetic data as a means to simultaneously protect personal information and enable meaningful public data usage.

Synthetic data is now recognized as the foundation for AI training, policy simulation, and public–private innovation.

🚀 DTS: A Non-Access Infrastructure Purpose-Built for Public Institutions

CUBIG DTS is uniquely designed around a non-access architecture where original data never leaves the institution. Instead of exporting raw data, the system learns only statistical patterns and generates entirely new synthetic datasets within a secure internal environment.

Advanced techniques such as Differential Privacy mathematically control re-identification risk, enabling agencies to validate safety before publishing or sharing data. DTS also includes built-in tools to measure statistical similarity, model performance, and privacy risks, providing quantitative verification essential for government compliance.

This ensures alignment with Korea’s data protection laws, the Data 3 Act, and even GDPR-equivalent requirements, making DTS suitable for handling high-risk public datasets.

💡 One Platform for Tables, Text, Images, and Time-Series

DTS supports a wide range of public sector data formats including tabular administrative data, civil complaint text, medical and industrial images, and sensor or log-based time-series data.

This unified pipeline allows institutions to build AI-ready data lakes, prepare datasets for the Public Data Portal, and support joint research across ministries and external organizations.

By converting sensitive datasets into synthetic versions, agencies can safely accelerate projects involving cross-agency collaboration, predictive modeling, and service innovation.

This enables a level of data usability that traditional anonymization cannot match, positioning synthetic data as the new standard for public sector AI.

🌍 Expanding AI-Ready Public Data Across Central and Local Government

CUBIG has already validated DTS through projects in finance, defense, and public agencies. Moving forward, the company plans to integrate DTS with SynTitan and SynLake to provide tailored AI-ready packages for ministries and local governments.

CEO Ho Bae emphasized that the goal of public data opening is shifting from simply releasing data to providing data that AI can immediately use for real-world impact. DTS empowers public institutions to share data safely with private and research organizations while advancing their own AI-driven administrative services.

CUBIG will continue expanding demonstrations, AI-ready dataset projects, and cross-ministry synthetic data collaborations to help shape the future of government innovation.

📎 Read more

Read the full article(Click)📰

#CUBIG #DTS #SyntheticData #AIReady #PublicData #DifferentialPrivacy #GovTech #SynTitan #SynLake #TrustedAI