Analysis Region: Gyeongsangbuk-do
Key Regions: Gyeongsangbuk-do Provincial Office, 22 Cities/Counties, 321 Eup/Myeon/Dongs, Invested/Funded Institutions, Regional Public Data and Administrative Services
Agenda: Will Generative AI, public data, spatial information, and administrative service platforms remain as individual tools, or will they be transformed into a local government operating system that proactively detects signals of change in the 22 cities and counties, enables responsible departments to report evidence and solutions back-report, and re-verifies policy outcomes?
Golden Time Type: Opportunity + Governance Integration Risk
Reference Date: August 28, 2026
Version: Regional AX Golden Time Intelligence v3.2
Entering 2026, Gyeongbuk Administrative AX moved beyond the tool introduction phase and entered a stage where a joint metropolitan-local platform and AI administrative services are being established simultaneously. Based on the "Gyeongsangbuk-do Intelligent Informatization Master Plan 2026–2030," finalized in December 2025, the province and 22 cities and counties will pursue approximately 720 informatization projects in 2026 with a total budget of 142.6 billion won . With the formalization of a joint response system at the meeting of provincial and city/county informatization department heads in February 2026, the unit of administrative digitalization has also shifted from the provincial government office to the metropolitan-local network. ( Gyeongbuk Provincial Government )
The AI operational infrastructure is already in operation. In April 2025, Gyeongbuk Province introduced closed-network-based generative AI document creation and summarization functions to **GBinPLUS+**, which is jointly used by the province and 22 cities and counties; at the time, approximately 3,000 provincial and municipal officials utilized the service. Given that GBinPLUS+ itself is a metropolitan-local platform that integrates administrative and public data—such as population, industry, and spatial information—to support policy decision-making, Gyeongbuk has begun to simultaneously secure both the Data Layer and the AI Layer necessary for Administrative AX. ( Gyeongbuk Provincial Government )
In July 2026, this structure was expanded to resident services. Selected for the Ministry of the Interior and Safety's Public Sector AI Service Support Project, a total of 1.2 billion won was invested to begin establishing AI settlement design, AI living support, small business support, pre-diagnosis for factory establishment, and AI civil complaint offices across 22 cities and counties. It connects the existing GBinPLUS+ with the pan-government AI common platform and applies a RAG structure that provides responses based on laws, guidelines, and public announcements. ( Gyeongsangbuk-do Sericulture and Insect Business Center )
A transformation is also being observed in spatial administration. In July 2026, Gyeongbuk Province completed the construction of an Intelligent Spatial Information Platform for joint use by the province and cities/counties, and introduced integrated 2D and 3D maps and AI spatial analysis. Starting in March, it began operating a function for surveying the status of provincial property using AI, a first in the nation. ( Gyeongbuk Provincial Government )
The foundational level of data governance is also not low. In the Ministry of the Interior and Safety's 2025 evaluation, Gyeongbuk received an " Excellent" rating for data-based administration for two consecutive years , and the operational status of public data provision improved from "Average" in the previous year to "Excellent." ( Gyeongbuk Provincial Government )
However, current evidence only demonstrates the readiness of the platform, AI capabilities, and data quality , and does not prove the outcomes that this asset actually detects policy issues first, compares the speed of change in 22 cities and counties, and shortens the administrative decision-making cycle.
The key gaps in the current Gyeongbuk administration AX are as follows.
- AI Document Creation ≠ AX Administration
- Data integration ≠ Change signal detection
- AI civil complaint service ≠ policy forecasting
- Spatial information platform ≠ Policy priority determination
- 720 IT projects ≠ One Government Operating System
Gyeongbuk's Golden Time lies in 2026–2028, when it establishes a common Runtime of Regional Data → Change Signal → Department in Charge → Evidence Analysis → Reverse Reporting → Policy Decision → Implementation → Outcome → Readjustment in 22 cities and counties, rather than the time to increase AI usage rates .
Gyeongbuk manages 22 cities and counties, 321 towns, townships, and neighborhoods, and approximately 2.5 million residents across a vast area of 18,424 square kilometers. Pohang, Gumi, and Gyeongsan, where population and industry are concentrated, coexist with county areas severely affected by extreme aging and population decline, all within the same metropolitan administrative system. As confirmed in No. 026, 15 cities and counties are areas of population decline, and 2 are areas of interest.
In this structure, identical policies do not operate at the same speed. Changes in businesses, transportation, and workforce in Gumi resulting from semiconductor investment, as well as issues regarding medical care, care services, and transportation in Yeongyang and Bonghwa, are difficult to manage using the same administrative indicators. The role of the provincial government must shift from managing the 22 cities and counties based on the same standards to comparing different speeds of change and detecting risks before they exceed a critical threshold .
The investment scale of 142.6 billion won across approximately 720 IT projects by 2026 demonstrates that the next bottleneck is not a lack of digital infrastructure, but rather how effectively distributed systems are integrated into a common architecture . ( Gyeongsangbuk-do Provincial Government )
Gyeongbuk already possesses a common platform for both metropolitan and local governments. GBinPLUS+ is jointly utilized by the province and 22 cities and counties, linking administrative and public data such as population, industry, and spatial information. Generative AI capabilities have also been introduced in a closed network environment. ( Gyeongbuk Provincial Government )
However, the primary uses identified in currently available data are document creation, summarization, and report support, as well as administrative data analysis. There is no evidence confirming that the process of automatic anomaly detection → automatic assignment to responsible department → comparison by city/county → generation of policy alternatives → reverse reporting → outcome verification operates as a standard process across the entire provincial administration.
While the Intelligent Spatial Information Platform has secured AI spatial analysis and a foundation for joint utilization by the province and cities/counties, the extent to which it is utilized as a common decision-making engine for policies regarding industrial location, disaster, welfare, transportation, and population decline remains to be verified. ( Gyeongsangbuk-do Provincial Government )
Currently, Readiness is assessed as having made significant progress in Digital Integration, but Decision Integration is in the verification stage.
The first structural change is that administrative AI is shifting from a productivity tool for individual public officials to a shared infrastructure spanning metropolitan and local governments . Given that 22 cities and counties use the same platform, GBinPLUS+ is more likely to evolve into a common administrative AI rather than a structure where generative AI is built separately for each city and county. ( Gyeongsangbuk-do Provincial Government )
The second point is the expansion of data types from administrative statistics to spatial, lifestyle, and industrial data. The Intelligent Spatial Information Platform, completed in 2026, was designed to enable the joint use of 2D and 3D spatial information and AI analysis by the provincial and municipal governments. The unit of policy analysis is shifting from tables and reports to spatial areas and living zones. ( Gyeongsangbuk-do Provincial Government )
The third point is the transition of resident AI from a simple chatbot to an agent supporting administrative procedures . Resident Care AI connects settlement, welfare, small business owners, factory establishment, and application form completion. It is a structure that shifts from a stage where AI merely provides information to a stage where it assists the administrative journey. ( Gyeongsangbuk-do Sericulture and Insect Business Center )
The fourth point is that AI capabilities are expanding from working-level staff to the decision-making tier. By 2026, the Gyeongbuk Human Resources Development Institute had expanded special AI training to include senior officials at Grade 4 and above, and is operating year-round basic and advanced training for provincial, city, and county officials at Grade 5 and below, as well as employees of public institutions. The policy-making stage has begun to be included as a target for AX. ( Southern Construction Office )
The Gyeongbuk Sovereign AI Basic Plan allocated four major strategies and 73 implementation tasks to each bureau and reviewed the progress in March 2026. The expansion of AX across the entire provincial administration, not just in industry, was included in the official implementation framework.
Generative AI has already been applied to GBinPLUS+ to support document creation and summarization for provincial and municipal officials. A structure has also been secured to reduce the risk of external leakage of administrative information by applying a closed-network-based standalone AI. ( Gyeongsangbuk-do Provincial Government )
On the residents' side, the Resident Care AI initiative was launched in 2026. Structured to allow residents of 22 cities and counties to comprehensively utilize administrative information regarding settlement, livelihood support, projects, and factory establishment, the application unit of AI administration has expanded from within the civil servant bureau to the residents' journey. ( Gyeongsangbuk-do Sericulture and Insect Business Center )
In the field of spatial information, the AI-based survey of provincial property and the Intelligent Spatial Information Platform are in operation, and in terms of policy ideas, a system has been launched to directly identify AI tasks from resident inconveniences through the 2026 Resident-Perceived AX Competition. ( Gyeongsangbuk-do Provincial Government )
The policy portfolio of Gyeongbuk Administration AX has already expanded from internal operations to data, space, resident services, and policy discovery. The remaining bottleneck is to combine these projects into a single Decision Runtime .
The relative strength of Gyeongbuk Administrative AX is that it already possesses a joint platform for metropolitan and local governments . While many local governments are adopting generative AI to assist with agency-specific tasks, GBinPLUS+ is structured so that Gyeongbuk Province and its 22 cities and counties share population, industrial, and spatial information along with AI functions. ( Gyeongbuk Provincial Government )
Furthermore, the results of receiving an "Excellent" rating in the 2025 Data-Based Administration evaluation for two consecutive years and the Public Data evaluation demonstrate that a certain level of data management and utilization readiness—a stage preceding AI—has been secured. ( Gyeongsangbuk-do Provincial Government )
On the other hand, evidence for Predictive Government , the higher level of Administrative AX, is still limited. There is no confirmed structure for publicly tracking when an indicator exceeds a threshold, which department responds, and how long it takes to modify the policy.
Therefore, Gyeongbuk's current position is in a transitional phase moving from an AI-enabled government to an intelligence-driven government .
In 2026, there are approximately 720 informatization projects in the province and 22 cities and counties , totaling 142.6 billion won . ( Gyeongsangbuk-do Provincial Government )
Approximately 3,000 provincial and city/county officials used the GBinPLUS+ AI document creation and summarization features when they were introduced in 2025. ( Gyeongbuk Provincial Government )
The Sovereign AI Basic Plan consists of 4 major strategies and 73 tasks .
Resident Care AI will establish five types of services across 22 cities and counties with a total investment of 1.2 billion won . ( Gyeongsangbuk-do Sericulture and Insect Business Center )
The Intelligent Spatial Information Platform was completed in 2026 on the premise of joint utilization by the province and cities/counties. ( Gyeongsangbuk-do Provincial Government )
These figures are evidence that Gyeongbuk is not remaining in the experimental stage of administrative AI.
However, current public KPIs are focused on the number of projects, platform construction, AI capabilities, training, and service development.
The numbers to be evaluated going forward are the number of signals detected, prediction accuracy, department connection time, policy decision time, reduction rate of recurring complaints, Outcome GAP by city/county, and time required for policy modification .
The biggest gap is between Integrated Data and Integrated Decision .
Even if GBinPLUS+ and the spatial information platform connect data, if a threshold is not defined as to what kind of change is a policy risk, the information remains on the Dashboard.
The second point lies between AI task assistance and organizational AX . While a user base of 3,000 generative AI users is a strength, a reduction in document creation time does not automatically speed up policy decision-making. ( Gyeongbuk Provincial Government )
The third is the outcome standardization gap between the province and the 22 cities and counties . Even with the use of a joint platform, if the policy performance of each city and county is managed using different indicators, comparison and learning at the metropolitan level are limited.
The fourth point concerns the 720 IT projects and the common architecture . If systems, data, and contractors for each project become entrenched individually, it becomes difficult for AI agents to utilize all administrative data. ( Gyeongsangbuk-do Provincial Government )
The fifth point concerns the relationship between AI services and the structure of responsibility . Once resident care AI begins providing guidance on laws, support policies, and factory establishment, standards regarding errors, update cycles, those responsible, and human review become directly linked to administrative trust. ( Gyeongsangbuk-do Sericulture and Insect Business Center )
Gyeongbuk is relatively well-equipped to transition existing platforms into a Government Intelligence Layer rather than creating a new single platform.
The data-based system can be handled by GBinPLUS+ and the data-based administration system.
The spatial layer can be handled by the intelligent spatial information platform.
The AI Layer can utilize Sovereign AI and the pan-government AI common platform.
Resident interaction can be handled by Resident Care AI and Moiso. Moiso utilizes Public MyData and blockchain to provide integrated services such as farmers' and fishermen's allowances, accompanying children to the hospital, and childcare. ( Gyeongsangbuk-do Sericulture and Insect Business Center )
What remains is Signal Governance .
The process of Indicator → Threshold → Responsible Department → Analysis Time → Reporting Responsibility → Policy Decision → Outcome must be institutionalized.
Similarly, regarding AI talent, the evidence verification capabilities of policymakers are more likely to become a bottleneck than those of development personnel. While the direction of expanding AI training to executives by 2026 is appropriate, actual changes in policy decisions must be verified through a separate runtime. ( Southern Construction Office )
Gyeongbuk's AI administration is shifting toward direct contact with residents. Moiso is integrating lifestyle-oriented administrative services such as farmers' and fishermen's allowances and care services, while Resident Care AI will expand its scope to include settlement, welfare, small business owners, and factory establishment starting in 2026. ( Gyeongsangbuk-do Sericulture and Insect Business Center )
However, resident sentiment cannot be determined by the number of platform subscribers or AI consultations.
Changes in administrative service search time, application error rates, revisits for civil complaints, factory establishment processing time, welfare omissions, and the non-beneficiary rate of small business support programs must be the outcomes.
Currently, the publicly available evidence does not compare these conversion effects across the 22 cities and counties. While Access Readiness of resident services is on the rise, Outcome Readiness shows a data gap.
The fact that 22 cities and counties utilize the same GBinPLUS+ is an important structural asset for Gyeongbuk. This is because it allows access to the same AI and data infrastructure regardless of regional size. ( Gyeongbuk Provincial Government )
However, platform access rights and analytical capabilities may not be the same. Large cities such as Pohang and Gumi differ from small counties like Yeongyang, Ulleung, and Bonghwa in terms of IT personnel, expertise, data volume, and administrative demands.
Therefore, Gyeongbuk should adopt a structure of regional Common AI, Common Data, and Common Security plus city/county Local Agents , rather than creating separate AI systems for each city and county.
An approach where metropolitan areas provide platforms and models, and local governments design their own problems, thresholds, and work workflows, is highly likely to reduce the AX gap.
The core timeline of Gyeongbuk Administrative AX lies in the fact that the 2026–2030 Basic Plan for Intelligent Informationization and the 73 Sovereign AI projects commenced simultaneously in 2026. ( Gyeongsangbuk-do Sericulture and Insect Business Center )
GBinPLUS+ Generative AI is already in operation, the Intelligent Spatial Information Platform was completed in 2026, and the construction of Resident Care AI also commenced this year. Now is the time to determine common data standards, APIs, and signal systems before each system grows independently. ( Gyeongsangbuk-do Provincial Government )
If the next two to three years are missed, 720 IT projects and new AI services could become entrenched with different data structures, operators, and performance indicators.
The irreversibility of administrative AX stems from the lock-up of workflows and data structures rather than the AI model itself.
12-1. Golden Time Application Case in Basic Local Governments ① — Andong City
Andong is the provincial capital and a region where administrative demands are rapidly diversifying due to the coexistence of the National Bio Industrial Complex, rural areas, an aging population, and new cities. It possesses the conditions to most directly link Gyeongbuk Province's common GBinPLUS+, spatial information, and resident care AI. ( Gyeongbuk Provincial Government )
The Readiness GAP lies not in adding individual services, but in managing industrial complex investment → corporate inflow → housing and transportation → population → schools and medical → administrative demand as a single signal chain. Andong's Golden Time is 2026–2028, when a growth-management type Government AX is established prior to the groundbreaking of the National Bio Industrial Complex .
12-2. Golden Time Application Cases in Basic Local Governments ② — Yeongyang-gun
Yeongyang is a region experiencing population decline characterized by a small population size and key administrative variables such as aging, transportation, medical services, and accessibility to daily living services. In this scenario, utilizing the Gyeongbuk common platform is more economical than building an independent AI system.
The Readiness GAP consists of a lack of data and a small administrative workforce. A structure is suitable in which regional data is combined with comparison models for adjacent cities and counties to automatically detect signals such as an increase in elderly single-person households, decreased transportation usage, risks to access medical care, and omitted welfare applications, while town and village officials conduct on-site verification. Yeongyang's Golden Time is the period to supplement administrative capabilities with Shared Government AI before workforce reductions lead to a decrease in administrative services .
The first loss is the entrenchment of silos across 720 IT projects . If project-specific data and functions are not integrated into a common structure, it will be difficult for future AI agents to connect the entire administration.
The second is the organization of the AI utilization gap among civil servants. If the use of generative AI remains a matter of individual capability, job knowledge is lost upon personnel transfers.
The third factor is the administrative disparity between cities and counties. Even if the platform is the same, the accumulation of differences in personnel utilization and analytical capabilities can widen the Government Intelligence gap between large cities and rural areas.
The fourth factor is the loss of initial outcome data. If baseline data for resident care AI and the intelligent spatial information platform is not collected now, it will be difficult to verify actual effects in two to three years.
Gyeongbuk does not need to operate its 22 cities and counties as one massive centralized administration.
A Distributed Government Intelligence structure can be established in which metropolitan areas provide common AI, data, security, and spatial information, cities and counties define local issues, and towns and villages verify field signals .
In this structure, even small-population archipelagos can utilize advanced administrative analysis without a large-scale AI development organization.
Performance is not measured by the number of AI uses, but by the process of shortening problem detection time → proactive response → reduced policy errors → improved resident outcomes.
It must be determined as.
Changes to observe | Things to do with AX | Policy decision | Verification indicators |
| 720 IT projects | Common Architecture · API | Coordination of duplication and individual construction | System reuse rate |
| GBinPLUS+ | Add Policy Signal Engine | Automatic classification of high-risk changes | Signal Hit Rate |
| 22 city/county data | Common Data Model | Comparison of identical indicators | Data linkage rate |
| population decline | Early Warning Agent | preemptive policy coordination | Net Mobility and Age Risk |
| Industrial investment | Growth Impact Agent | Infrastructure pre-installation | Investment → Administrative Demand Forecast |
| Resident Care AI | Journey Outcome Tracking | Service improvement | Search and processing time |
| Spatial information | AI Spatial Risk Analysis | Facility, transportation, and welfare placement | Access time |
| Repeated complaints | Pattern Detection | Removal of structural cause | Recurrence rate |
| Civil Servant AI | Shared Workflow·Agent | Standardization of best practices | Reuse rate |
| Policy outcomes | Budget-to-Outcome Connection | Continue · Edit · End | Outcome Achievement Rate |
Government AX Runtime
Local Data → Change Signal → AI Analysis → Automatic Connection to Responsible Department → Evidence Verification → Solution Back-reporting → Policy Decision → Implementation → Outcome → Comparison of 22 Cities/Counties → Policy Readjustment
Opportunity + Governance Integration Risk
Gyeongbuk's administrative AX basic Readiness is high.
GBinPLUS+, jointly used by 22 cities and counties, exists, and generative AI has already been applied. ( Gyeongsangbuk-do Provincial Government )
The 2026–2030 Basic Plan for Intelligent Informationization and approximately 720 informatization projects worth 142.6 billion won are being implemented simultaneously. ( Gyeongsangbuk-do Provincial Government )
Both data-based administration and public data evaluations received an excellent rating. ( Gyeongsangbuk-do Provincial Government )
With the addition of the Intelligent Spatial Information Platform and Resident Care AI, the major layers of data, AI, space, and resident services are also being established. ( Gyeongsangbuk-do Provincial Government )
However, the common runtime for the entire process from Signal → Back-reporting → Policy Decision → Outcome is not currently confirmed in the publicly available evidence.
The danger for Gyeongbuk is not a lack of AI.
As AI and data increase, there is a risk that administrative processes will become disconnected.
Therefore, Golden Time is determined by Opportunity + Governance Integration Risk .
- Number of actual active users in provinces and cities/counties
- Generative AI Usage Rates by Department and City/County
- AI document creation time savings
- Number of Common AI Workflow Registrations
- Workflow Reuse Rate
- AI Result Error/Correction Rate
- AI Answer Source Verification Rate
- Application rate of common APIs for 720 IT projects
- Number of duplicate system reductions
- Number of common datasets for provinces and cities/counties
- Real-time/Automatically Linked Data Ratio
- Number of departments utilizing the intelligent spatial information platform
- Number of AI spatial analysis policy reflections
- Number of actual users of Resident Care AI
- Changes in AI complaint processing time
- Change from Pre-factory Establishment Diagnosis to Actual Processing Time
- Reduction rate of omissions in welfare and care services
- Detection and recurrence rate of repeated complaints
- Number of policy risks detected by AI
- Signal → Connection time to responsible department
- Signal → Time required for policy decision
- Number of joint response cases between cities and counties
- Error between policy predictions and actual results
- Government AX utilization gap in 22 cities and counties
- Changes in Outcome After Policy Revision
Data GAP: While Gyeongbuk discloses the performance of its respective platforms for AI, public data, spatial information, and administrative services, common Government Intelligence data across the province and its 22 cities and counties, which connects signal discovery → policy intervention → resident outcomes, has not yet been verified.
Runtime Chain
Regional/Administrative Data → Change Detection → AI Signal → Connecting Departments/Cities/Counties → Evidence Verification → Back-reporting → Policy Decision → Implementation → Outcome → Regional Comparison → Learning/Redesign
The strongest evidence for Gyeongbuk Administrative AX is that the province and 22 cities and counties already share a single GBinPLUS+ platform and generative AI environment . At the time of its introduction in 2025, approximately 3,000 public officials were using the platform, and it integrates metropolitan and local administrative data, including population, industry, and spatial information. ( Gyeongbuk Provincial Government )
In 2026, building upon this foundation, approximately 720 provincial and municipal informatization projects worth 142.6 billion won, along with the Intelligent Spatial Information Platform and Resident Care AI, were simultaneously added. The evaluation of data-based administration has also been rated excellent for two consecutive years. Therefore, it is difficult to classify Gyeongbuk as a region where the transition is difficult due to a lack of basic infrastructure for Administrative AX. ( Gyeongbuk Provincial Government )
Conversely, the point where evidence is currently lacking is whether AI has actually changed the decision-making process . While it is confirmed that AI drafts documents and guides residents on policies, a standardized runtime—in which it first detects local changes, transmits them to the relevant department, and then relearns the results after implementation—has not been publicly verified.
There is only one structural insight. Gyeongbuk’s administrative AX competitiveness is determined not by the number of AI functions, but by how quickly a single Intelligence Network can compare and assess the varying speeds of change across its 22 cities and counties. Due to Gyeongbuk’s vast area and numerous regions experiencing population decline, the limitations of an administrative model where humans observe every site with uniform density are likely to become apparent more rapidly than in the Seoul metropolitan area.
Golden Time Thesis — The golden time for Gyeongsangbuk-do’s Administrative AX is not the time to add more generative AI capabilities to GBinPLUS+ or increase the number of informatization projects to 720. The key lies in whether, within the next two to three years, we can transition into a local government operating system that can detect the differing speeds of change among 2.5 million provincial residents and 22 cities and counties by connecting the already secured joint metropolitan-local platforms, data, spatial information, and resident AI through a process of Regional Change → Signal → Responsible Department → Evidence → Reverse Reporting of Solutions → Policy Decision → Implementation → Outcome → Comparison of 22 Cities and Counties. If this connection is missed, Gyeongsangbuk-do may remain at the level of simply performing existing administrative processes faster, despite possessing nationally advanced AI administrative tools. If the connection is successful, the vast area and dispersed regional structure itself can become the largest Distributed Government Testbed for verifying local government AX in South Korea.
Version | Reference Date/Revision Date | Major changes |
| v1.0 | 2026.08.28 | The Gyeongbuk Provincial Administration AX Readiness was analyzed for the first time based on Regional AX Golden Time Intelligence v3.2. The analysis cross-verified the 2026–2030 Basic Plan for Intelligent Informationization, approximately 720 informatization projects totaling about 142.6 billion KRW across the province and 22 cities/counties, the GBinPLUS+ integrated metropolitan-local platform and a user base for generative AI with about 3,000 people, the four major strategies and 73 implementation tasks for Sovereign AI, two consecutive years of excellence in data-based administration and excellent public data evaluations, the 2026 Intelligent Spatial Information Platform, 1.2 billion KRW in 5 types of resident care AI services across 22 cities/counties, and AI training for public officials. The Government AX Runtime of Data Layer → AI Layer → Signal Layer → Reverse Reporting → Decision → Outcome was presented, and Andong and Yeongyang were analyzed as application cases for local governments; the Golden Time was determined to be Opportunity + Governance Integration Risk. |
