Samsung SDS Expands Brity Works Across Korean Ministries
Samsung SDS announced on August 13 that nine South Korean central government ministries have adopted its AI collaboration platform, Brity Works, with phased service rollout scheduled through August. DigitalToday reports that Brity Works is the official collaboration tool for the government's OnAI platform, which the Ministry of the Interior and Safety intends to extend to more than 40 central agencies in the second half of 2026.
Samsung SDS announced on August 13 that nine South Korean central government ministries have adopted Brity Works, its generative AI-enabled collaboration platform. The service is scheduled for phased rollout through August, according to DigitalToday and Chosun Biz.
The adopters include the Ministry of Patriots and Veterans Affairs, Ministry of Trade, Industry and Energy, Ministry of Employment and Labor, Financial Services Commission, and Overseas Koreans Agency. The deployment follows earlier adoption by the Ministry of Finance and Economy and the Ministry of Planning and Budget, DigitalToday reported.
Brity Works within the OnAI program
Brity Works serves as the official collaboration tool for OnAI, a generative AI-based intelligent work-management platform promoted by the Ministry of the Interior and Safety. Chosun Biz reported that the ministry has announced an expansion of OnAI to more than 40 central government agencies in the second half of 2026.
According to DigitalToday, Brity Works combines AI search, knowledge management, document-writing support, messaging, videoconferencing, and file sharing. Its reported public-sector workflows include searching large administrative data collections, producing organized meeting-minute summaries, and managing follow-up work.
Maeil Business Newspaper reported that Samsung SDS began official service in April after pilot deployments at major ministries from November 2025 and user evaluations. The publication also reported mobile functions for reporting, approvals, videoconferencing, and AI-assisted meeting summaries.
Security and deployment considerations
DigitalToday reported that the platform operates on Samsung SDS security technology and a public-private partnership cloud environment that received a top-level "high" security rating from South Korea's National Intelligence Service. The retrieved report does not specify the underlying foundation model, retrieval architecture, data-retention terms, or controls governing access to agency data.
For AI and data practitioners, the deployment illustrates a public-sector pattern in which generative AI enters production through existing collaboration workflows rather than standalone chat interfaces. At scale, comparable rollouts typically make identity management, retrieval permissions, audit trails, evaluation of summarization quality, and mobile-device controls central implementation concerns. Those operational details have not been publicly described for this deployment.
The reported expansion of OnAI to more than 40 agencies makes interoperability and governance relevant questions for observers, particularly whether agencies can maintain distinct data-access boundaries while sharing a common collaboration layer.
Key Points
- 1Nine Korean ministries adopted Brity Works, extending Samsung SDS's AI collaboration deployment beyond its initial finance and planning ministry users.
- 2Brity Works combines retrieval, document assistance, meeting summaries, messaging, and videoconferencing, embedding generative AI in routine government collaboration workflows.
- 3Comparable cross-agency AI deployments commonly depend on rigorous identity, retrieval-permission, audit, and data-governance controls as adoption broadens.
Scoring Rationale
This is a notable production deployment of generative AI collaboration tooling across South Korean central government ministries, with a potentially larger OnAI expansion reported for later in 2026. It matters to practitioners working on secure, governed enterprise AI deployments, although the available reporting provides limited technical detail about models, architecture, or evaluation.
Sources
Public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
