알림 목록

Analysis Region: 18 cities and counties in Gyeongsangnam-do
Key Industries:  Machinery, Defense, Shipbuilding, Aerospace, Automotive Parts, Smart Manufacturing
Agenda: The spread of AI education and the actual connection of the Manufacturing AX talent supply chain
Golden Time Type: Educational Infrastructure Opportunity + Industrial Talent Pipeline Risk
Version: Regional AX Golden Time Intelligence v3.2 Enhanced Criteria Reference
Date: 2026.08.28
 

The Gyeongnam Office of Education has been operating 'iTalkTalk,' an AI and big data-based education platform, since 2021, and as of 2025, it has accumulated approximately 300,000 educational content items and over 7TB of learning data. By April 2026, 887 out of 998 elementary, middle, and high schools in the province had registered for the linked AI math service, recording an 89% school adoption rate. While AI-based teaching and learning infrastructure has expanded significantly at the metropolitan office level, there is no confirmed evidence linking this data to job competencies in smart manufacturing, aerospace, shipbuilding, and defense. If subject learning data and industrial job data remain separated between 2026 and 2028, the supply of Manufacturing AX talent will stagnate separately, even if AI education utilization rates rise. Although  Gyeongnam education's AI learning infrastructure has entered the leading tier, its Manufacturing AX talent system has not yet reached the same stage. ( Electronic Times ,  Chosun Edu )

The Gyeongnam Office of Education designated aviation, smart manufacturing, shipbuilding, future vehicles, biomedical science, elevators, and nano-semiconductors as strategic fields for vocational education innovation districts, and over the past three years, restructured 18 departments across 14 vocational high schools into fields such as AI, smart manufacturing, AIoT, and aviation. In 2025, Yangsan Artificial Intelligence High School opened, featuring four departments: AI Convergence Factory, AI Automatic Control System, AI Content, and Bio-Food. With 302 applicants for the first 126 freshman spots, the school recorded a competition ratio of 2.4 to 1. While the names of schools and departments have shifted toward the direction of industrial transformation, the period of accumulation for results linking actual corporate tasks, project execution capabilities, post-graduation manufacturing employment, and regional retention is still short. The period from 2026 to 2028 marks the first phase where the names of the restructured departments are verified by actual job competencies.  Gyeongnam's Manufacturing AX talent education is assessed as being in the pre-verification stage, not the pre-dissemination stage. ( Gyeongnam Office of Education ,  Yonhap News )

In 2025, there were 3,139 graduates from vocational high schools in Gyeongnam, with 895 employed and an employment rate of 55.9%. While this was 0.7 percentage points higher than the national rate of 55.2%, it fell by 3.6 percentage points from the previous year. The university entrance rate also dropped to 44.9%, down 2.8 percentage points from the previous year. The structure is such that while departments focused on future industries in vocational high schools have expanded, graduates' entry into the labor market has simultaneously weakened. There is a lack of regional evidence that simultaneously illustrates the industry, company size, job function, wages, one- and three-year retention rates, and persistence in Gyeongnam for employed graduates. If the weakening of employment pathways outpaces curriculum restructuring over the next two to three years, the Manufacturing AX talent policy will fail to descend from the student recruitment stage to the industrial supply chain.  Currently, the "Golden Time" is Educational Infrastructure Opportunity + Industrial Talent Pipeline Risk. ( Reported by the Ministry of Education and  Gyeongnam Domin Ilbo )

Gyeongnam simultaneously possesses machinery, defense, and nuclear power industries in Changwon; shipbuilding in Geoje; aerospace in Sacheon; and automotive parts, machinery, and smart manufacturing in Gimhae and Yangsan. The Changwon National Industrial Complex alone is projected to generate 62.223 trillion won in production, $18.429 billion in exports, house 3,216 companies, and employ 120,289 people in 2024. While the industrial structure has shifted from a single manufacturing technology to a complex technology structure combining machinery, electronics, software, data, quality, and supply chains, school education remains fragmented into subject areas, departments, and certifications. There is no data comparing actual tasks by industry with student competencies using the same classification system. If the curriculum fails to reflect job requirements while large-scale orders transition into production between 2026 and 2028, the time lag between the industrial boom and the supply of new talent will widen.  Although Gyeongnam has sufficient industrial sites to demonstrate Manufacturing AX, the speed at which this is translated into the curriculum remains unconfirmed. ( Yonhap News )

The Gyeongnam Vocational Education Innovation District has expanded its fields to include aviation, smart manufacturing, shipbuilding, future vehicles and biomedical sciences, elevators, and nano-semiconductors. This serves as evidence that the external scope of vocational education's response has begun to align with that of local industries. However, the number of students by field, participation rates in industry projects, employment rates in related roles, and retention rates at partner companies are not disclosed in an integrated manner. Even if the number of educational fields matches the number of industrial fields, the pathway a student takes from admission to employment remains unverified. If, even after two to three years, project performance by field remains limited to the number of participants and partner institutions, the effectiveness of talent supply cannot be assessed. While  Gyeongnam vocational education has secured the capability to respond to the industrial portfolio, the completeness of its talent pipeline has not been confirmed. ( Gyeongnam Office of Education )

AI education in Gyeongnam proceeds along two axes: learning platforms in general elementary, middle, and high schools, and industrial education in vocational high schools. iTalkTalk accumulates extensive student learning data, while vocational high schools are responsible for industry-specific practical training and employment. However, there is no connecting indicator between these two axes that links learning competencies in mathematics, science, and information technology to capabilities in manufacturing design, automatic control, quality analysis, and predictive maintenance. If platform learning and vocational education operate as separate systems for two to three years, the entry pathways for general high school students into manufacturing AX and the foundational AI competencies of vocational high school students will become simultaneously unclear.  Although Gyeongnam education possesses both an AI learning system and a manufacturing talent system, it lacks evidence that they are connected as a single educational supply chain.

Yangsan AI High School opened in March 2025 with the establishment of the AI ​​Convergence Factory, AI Automatic Control Systems, AI Content, and Bio-Food departments. It was equipped with a capacity of 378 students across 21 classes and practical training facilities that simulate actual industrial sites. Although the educational infrastructure for an AI-specialized school has been formed, the first graduating class has not yet been produced; therefore, performance data regarding manufacturing employment rates, job suitability, corporate evaluations, and regional retention does not exist. The period from 2026 to 2028 marks the time when the first verification data regarding the completion of educational facilities and talent outcomes will begin to accumulate.  Yangsan AI High School is assessed not as a success model, but as a pioneering model where performance verification has begun. ( Gyeongnam Office of Education )

Over the past three years, 18 departments across 14 vocational high schools in Gyeongnam Province have been reorganized into fields such as AI, smart manufacturing, AIoT, and aviation; furthermore, in 2026, five departments at five schools were selected for the restructuring project. While the speed of department reorganization is confirmed, details regarding faculty major conversions, the field consistency of industrial equipment, access to corporate data, and changes in project evaluation methods are not disclosed on a school-by-school basis. A gap is observed between changes in department names and changes in curriculum tasks. If reorganized departments remain at the level of merely adding AI subjects to existing curricula for the next two to three years, the convergence capabilities for Manufacturing AX will not be formed.  Currently, department restructuring is in the stage of quantitative diffusion, and it is impossible to assess the level of industry synchronization of educational content on a school-by-school basis. ( Yonhap News )

In 2025, the employment rate for vocational high school graduates nationwide was 55.2%, the proportion of graduates employed by companies with 300 or more employees was 36.3%, the first-year retention rate was 83.1%, and the second-year retention rate was 68.2%. A national statistical structure has been established that does not judge the quality of employment based solely on the employment rate, but rather considers company size and the retention rate together. In Gyeongnam, while the regional employment rate of 55.9% is confirmed, the proportion of employment in manufacturing, the proportion of employment in companies with 300 or more employees, the retention rate, and the provincial employment rate are not disclosed in a single table. Even for the periods 2026–2028, if only the total employment rate is presented, it cannot be verified whether the reorganization of future industry departments has led to entry into high-quality manufacturing.  The segmentation of outcome evidence regarding vocational education in Gyeongnam lags behind the policy diffusion. ( Ministry of Education )

The AI ​​transformation in the manufacturing sector is extending beyond equipment automation to encompass design, process planning, quality inspection, predictive maintenance, production scheduling, supply chain management, and safety management. The boundaries between production and technical roles are blurring, creating a structure where mechanical, electrical, and electronic knowledge is combined with data interpretation and AI verification capabilities. While the restructuring of vocational curriculums in Gyeongnam reflects this direction of change, the alignment between school practical training and the data, software, and equipment used in the industrial field remains undisclosed. If corporate adoption of AX outpaces school equipment upgrades and teacher retraining over the next two to three years, the skills acquired by graduates will already lag behind on-site standards.  The time-sensitive risk facing manufacturing AX talent stems not from a lack of training, but from the shortened shelf life of educational content.

The Bank of Korea analyzed that of the 285,000 youth jobs lost between 2022 and 2026, 268,000 (94%) occurred in sectors highly exposed to AI. While youth employment in information services decreased by 31.4%, computer programming by 16.6%, and professional services by 11.6%, employment for those in their 50s in the same sectors increased. AI has manifested as a structural change that complements experienced workers while compressing tasks previously handled by entry-level professionals. Education in Gyeongnam lacks the standing evidence to distinguish between disappearing and new tasks within entry-level manufacturing roles. Unless entry-level roles are redesigned within two to three years, students will be unable to secure their first career experience even if they learn AI.  The risk to manufacturing education in Gyeongnam lies not in failing to teach AI, but in the inability to identify the initial roles that will remain after the introduction of AI. ( Bank of Korea )

The proliferation of generative and industrial AI places greater emphasis on problem definition, data verification, and on-site judgment than on knowledge acquisition. While iTalkTalk performs individual student learning analysis and content recommendations, it is not connected to project data dealing with defects, breakdowns, delivery times, and safety issues in manufacturing sites. A verification gap exists between the proficiency to use educational AI and the problem-solving ability to use industrial AI. As the use of AI tools becomes widespread between 2026 and 2028, the distinctiveness of simple application training will disappear, leaving only the experience of solving actual field problems.  Gyeongnam education has entered a stage where its competitiveness is judged not by AI usage rates, but by its history of solving manufacturing problems.

iTalkTalk operates as an AI and big data teaching and learning platform that connects curriculum, instruction, assessment, and records, possessing approximately 300,000 pieces of content and over 7TB of learning data. The projected 89% registration rate for AI Mathematics schools by 2026 serves as evidence that the use of AI in general subjects has expanded beyond pilot schools to a broader regional level. However, the connection between the mathematics, science, and information competencies accumulated on the platform and their subsequent impact on vocational high school major selection, industrial projects, and corporate recruitment has not been confirmed. If educational data remains confined within the curriculum for two to three years, it cannot be utilized for predicting manufacturing AX talent. While  iTalkTalk has expanded as an AI education infrastructure, its expansion into an industrial talent infrastructure remains unconfirmed.

Vocational Education Innovation Districts connect local governments, education offices, schools, universities, and companies across regional strategic industries such as aviation, smart manufacturing, and shipbuilding. Cooperation committees and support centers exist, and the cultivation of industry-specific technical talent is explicitly mandated. However, the actual number of tasks provided by partner companies, the class participation time of corporate experts, the field application rate of student projects, and the employment conversion rate are not disclosed in an integrated manner. A gap remains between the connection of participating institutions and the connection of educational tasks. If the cooperation system remains centered on meetings, agreements, and field trips from 2026 to 2028, it will fail to reflect the speed of industrial change in the curriculum.  Currently, while the organizational structure for industry-academia collaboration exists, the intensity of its runtime cannot be assessed. ( Gyeongnam Vocational Education Innovation District Support Center )

The Gyeongnam-type RISE has reorganized its university support system with the goal of fostering talent tailored to the demands of key regional industries and resident local talent, presenting a support plan of up to 340 billion won for six universities from 2024 to 2028. While both high school vocational education and the cultivation of high-level university talent aim for regional industries, no common performance indicators connecting high schools to junior colleges/universities and then to enterprises have been identified. There is a gap in the connectivity of learning, further education, and employment data between the talent policies of the Office of Education and the Provincial Government. If RISE and the Vocational Education Innovation Districts maintain separate performance systems for two to three years, talent projects for the same industries will become entrenched in different supply chains.  Although Gyeongnam's talent policy includes projects for specific educational stages, the lifelong pipeline remains incomplete. ( Gyeongsangnam-do Youth Information Platform )

Of the 59,661 vocational high school graduates nationwide in 2025, 15,296 were employed and 29,373 pursued further education. While the employment rate stood at 55.2%, the college admission rate rose to 49.2%, indicating that vocational high schools are shifting from an education system focused solely on immediate employment to a path where both employment and further education coexist. Gyeongnam recorded an employment rate of 55.9%, slightly higher than the national average; however, the decline from the previous year was 3.6 percentage points, which was greater than the national average drop of 0.1 percentage points. Despite being a region with a high manufacturing intensity, there is no evidence by industry or academic discipline explaining why vocational high school employment has weakened more significantly. If this trend persists for two to three years, the premise that manufacturing agglomeration leads to employment superiority in vocational education will collapse.  Gyeongnam does not hold a stable advantage in labor market outcomes for vocational education relative to its manufacturing base. ( Korea Educational Development Institute )

The proportion of vocational high school graduates nationwide employed by companies with 300 or more employees rose from 22.5% in 2021 to 36.3% in 2025, and the second-year retention employment rate also improved to 68.2%. This structure indicates that the benchmark for evaluating vocational education outcomes has shifted from employment status to company size and employment sustainability. While a total of 895 employed graduates in Gyeongnam are confirmed, the distribution of large, medium, and small enterprises, as well as the disparity between prime contractors and subcontractors, are not separated in the publicly available data. Even if graduates from restructured departments are produced between 2026 and 2028, if the quality of employment is not disclosed, the region's position relative to the national average will be misinterpreted solely based on the total employment rate.  Gyeongnam education lags behind the transition of its outcome indicators toward future industries compared to the transition of its curriculum to future sectors.

Gyeongnam operates its own AI teaching and learning platform at the city/province level and has secured an 89% school enrollment rate. While the diffusion level of general AI education infrastructure is relatively high, broad regional standards for manufacturing-specialized projects, industry data, digital twin practice, and joint corporate evaluations have not been identified. A gap exists between the leadership in general-purpose AI education and the leadership in Manufacturing AX education. If other regions directly link industry-centered projects with recruitment over the next two to three years, Gyeongnam's platform leadership will not extend to leadership in industrial talent.  The national strength of Gyeongnam education lies in its AI learning foundation, and the results of Manufacturing AX are not yet at a stage where they can be compared.

While 887 out of 998 schools in the province have registered for AI math services, only 14 vocational high schools and 18 departments have been restructured around future industries over the past three years. Although general-purpose AI learning has spread to most schools, industry-specialized education remains concentrated in a limited number of schools and departments. The rate at which AI learning capabilities accumulated in general high and middle schools are transferred to vocational education and manufacturing career paths remains unconfirmed. Without a career connection between the two systems for the next two to three years, the expansion of AI education and the shortage of manufacturing talent will persist simultaneously.  In Gyeongnam, while the number of students exposed to AI has increased rapidly, it is impossible to determine the increase in students entering the manufacturing AX career path.

In its first year, Yangsan AI High School received 302 applications for 126 available spots, while the total number of vocational high school graduates in Gyeongnam was 3,139, with 895 employed. Although demand for specific AI-specialized schools is high, the employment performance of vocational high schools as a whole has declined. The structure is such that the interest of students and parents in future industry education does not move at the same pace as the labor market's absorption capacity after graduation. There is a lack of mid-to-long-term evidence linking school application competition rates with industrial hiring demand. If applications for AI-specialized departments increase between 2026 and 2028 but the number of high-quality first jobs does not, departmental preference will shift back to university admission or relocation outside the province. While  expectations at the admission stage have been confirmed, the rewards at the employment stage remain unverified.

The 120,289 employees at the Changwon Industrial Complex and the 895 employed vocational high school graduates in Gyeongnam highlight the disparity in scale between total industrial employment and the supply of new high school graduates. Although these two figures cannot be converted into a direct employment ratio due to differences in survey subjects and scope, a data gap is confirmed, indicating a failure to link the retirement outlook by age group within the industrial complex with the entry of new vocational high school graduates. If the retirement of skilled Baby Boomers intensifies in two to three years, a shortage of replacement personnel by job function will appear before total employment statistics.  Gyeongnam measures the current scale of employment but fails to assess the potential for filling the ranks of the next generation of skilled workers.

The biggest gap lies in the separation between education for learning AI and education for solving manufacturing problems using AI . iTalkTalk accumulates learning achievement and subject data, vocational high schools manage practical training, qualifications, and employment, and companies possess production, quality, and equipment data. These three types of data do not link students' competencies with the results of solving industrial problems as identical evidence. As AI utilization becomes a universal competency between 2026 and 2028, training in simple tool usage will lose competitiveness, and only a history of solving actual manufacturing problems will remain as a hiring signal. While  Gyeongnam Education is highly prepared to create a large number of AI users, it is low in its readiness to verify manufacturing AX problem solvers.

The second gap is the speed of faculty transition to Industry AX. Over the past three years, 18 departments have been reorganized around new industries, and five additional departments were selected for 2026. However, data proportional to the speed of department reorganization—such as faculty dispatch to industry, data and AI retraining hours, industrial certifications, and joint project performance—are not disclosed. The structure requires the long-term accumulation of teachers' job knowledge, even if practical training equipment and department names change in the short term. If the faculty competency gap does not narrow within two to three years, the latest equipment will be used for traditional training methods.  The bottleneck for manufacturing AX education in Gyeongnam is likely to be the faculty's capacity to translate the curriculum into the field, rather than the number of students.

The third gap lies in post-graduation tracking. While national statistics present employment and retention rates, long-term retention outcomes by industry, job function, and region in Gyeongnam are not disclosed. To assess the success of department restructuring, one must examine one- and three-year tenure, wage increases, job transfers, and retention in Gyeongnam, rather than simply focusing on employment status; however, currently available evidence is limited to total employment rates. If the first graduation cohort from 2026 to 2028 is missed, it becomes difficult to separate the educational effects of individual departments from subsequent policy changes. The  "Golden Time" for tracking outcomes is not after graduates have been produced, but is currently in progress, starting from the moment the first cohort enrolled.

iTalkTalk has accumulated approximately 300,000 pieces of content and over 7TB of learning data, and 89% of schools in the province have registered for AI Math. Educational data infrastructure and school accessibility have expanded to the wider regional level. However, the scope of linkage with industry data regarding tasks, job functions, recruitment, and tenure remains unconfirmed. A gap persists between the scale of educational data and its ability to identify industrial talent. While the data grows if only learning information is accumulated over two to three years, the accuracy of manufacturing talent prediction does not improve.  The readiness of the educational infrastructure is assessed as high, while the readiness for linking with manufacturing data is assessed as low.

Yangsan AI High School and 18 restructured departments across 14 schools have formed educational hubs in the fields of smart manufacturing, automatic control, AIoT, and aviation. While the number of hubs has increased, the consistency of practical training equipment with industrial sites, faculty expertise, and the intensity of joint corporate classes vary by school, and there is no integrated evidence. A gap is identified between the establishment of schools and educational quality standards. If the differences between schools accumulate between 2026 and 2028, competency levels will differ under the same AI and smart designations.  Although the expansion of hubs is underway, the broad standardization of educational quality remains unconfirmed.

Vocational Education Innovation Districts and RISE each possess institutional connections with education offices, provincial governments, schools, universities, and enterprises. While a cooperative structure exists, a common identification system and joint performance indicators linking high school learning history, university education, and corporate recruitment are not identified. A gap remains between institutional connections and data connections. If separate governance is maintained for two to three years, the same students and companies will be counted as participants in multiple projects.  Institutional readiness is assessed as upper-middle, while integrated runtime readiness is assessed as low.

In 2025, out of 3,139 graduates from vocational high schools in Gyeongnam, 895 were employed, resulting in an employment rate of 55.9%. While future industry education policies were being expanded, the employment rate declined by 3.6 percentage points compared to the previous year. Although policy effects cannot be causally determined based solely on single-year figures, evidence confirms that policy diffusion and labor market outcomes did not move in the same direction. Due to the lack of decomposition data by industry, major, and firm size, the causes of the decline cannot be identified as manufacturing stagnation, further education, or job mismatches. Without this decomposition for two to three years, it is also impossible to determine whether the changes in academic departments have spread to the field.  What has benefited students are specific majors and educational opportunities, while the expansion of an employment premium remains unconfirmed.

The Gyeongnam Vocational Education Innovation District operates a cooperation committee involving companies, institutions, and schools in fields such as aviation and smart manufacturing. While institutional channels for corporate participation exist, the amount of time students spend handling actual corporate data, the evaluation ratio by corporate experts, and the scale of conversion to employment after projects are not disclosed. There is no distinction between a company being registered as a collaborating institution and participating as a co-producer of the curriculum. If this gap persists from 2026 to 2028, the outward scale of industry-academia collaboration may expand, but the on-the-job adaptation period for new employees, as perceived by companies, will not decrease.  While agreements reaching down to the companies can be verified, the training provided at the level of corporate tasks cannot be assessed.

The 2.4:1 competition ratio for the first admissions of Yangsan AI High School demonstrates the demand for AI-specialized education among students and parents. However, there is no evidence yet that the expectations associated with choosing this school will translate into employment and long-term retention at manufacturing companies in Gyeongnam three years from now. A time lag gap exists between admission competition and labor market rewards. Unless the quality of employment is verified by 2027–2028, when the first graduating cohort is formed, initial expectations will remain merely about the school's brand.  While the expansion of educational demand has been confirmed, the expansion of the supply of industrial talent has not yet occurred.

Changwon is connected to machinery, defense, and nuclear power; Sacheon to aerospace; Geoje to shipbuilding; and Gimhae and Yangsan to machinery, automotive parts, and smart manufacturing. Vocational education is also specialized around schools near industrial hubs, but mobility evidence connecting students' residences, schools, training companies, and employers is not disclosed. The structure places the burden of commuting, dormitory, and relocation costs on students in rural areas without industries. If the reorganization centered on hub schools continues for two to three years, disparities in educational accessibility will become entrenched as disparities in opportunities to enter the industry.  The spatial gap in Gyeongnam Manufacturing AX education is widening due to the distance to industrial sites rather than the number of schools.

Yangsan AI High School is located in the industrial and population-dense region of eastern Gyeongsangnam-do, while Changwon Mechanical Engineering High School is situated on the central manufacturing axis. In contrast, the reorganization of the AI ​​Energy Department at Hamyang Jeil High School represents a shift aimed at establishing a new industrial education hub in the western inland region. Although the establishment of departments is becoming dispersed, the hiring demands of local companies and the density of equipment, specialized faculty, and project partners are not uniform. A gap remains between the regional placement of schools and the regional placement of industrial ecosystems. If students receive their education locally but find employment elsewhere between 2026 and 2028, the dispersion of departments will not lead to the retention of local talent. While  the expansion of educational hubs in western Gyeongsangnam-do has been confirmed, its capacity to absorb industry remains unverified.

Geoje and Sacheon possess national-level anchor companies but exhibit a high degree of dependence on a single industry, whereas Changwon, Gimhae, and Yangsan have relatively diverse corporate groups. There is no confirmed evidence of students transitioning to different jobs within the province following economic fluctuations in a specific industry. A gap exists between city-specific specialized education and the province-wide career mobility network. If education remains fixed as tailored to a single industry for two to three years, students' skills move out of the region along with the industry during downturns. While  industry-specific specialized education is underway in Gyeongnam, a common competency framework that enables transition within the province remains unconfirmed.

Over the past three years, 18 departments across 14 schools have been reorganized around future industries, and Yangsan Artificial Intelligence High School selected its first students in 2025. In 2026, five departments across five schools were again selected for the restructuring project. Currently, we are in a stage where investment, teacher training, curriculum, and corporate collaboration are not yet solidified as the initial graduation outcomes. Unless job roles, projects, employment, and residual evidence are connected from the beginning, it is difficult to restore the effectiveness of each curriculum after graduation. The period from 2026 to 2028 is the only baseline segment where changes in the initial cohort are observed.  It is not currently deemed the time to create more departments, but rather the time to verify the industrial adaptability of the departments already established.

Ninety-four percent of the nationwide decline in youth jobs in AI-exposed sectors occurred within these industries. Even in the manufacturing sector, entry-level tasks such as design assistance, documentation, inspection recording, and basic analysis are directly impacted by AI and automation. Students trained by restructuring departments in Gyeongnam enter the post-automation labor market immediately upon graduation. While the gap between existing job titles and actual remaining tasks is widening, there is no evidence to update this gap on a semi-annual basis. There is a risk that the job roles designed for students at the time of admission will have been reduced when the first graduating class is produced in two to three years. The end of the "Golden Time" is not the completion of school reorganization, but the moment the curriculum lags behind industrial change by a generation.

The expansion of orders in Gyeongnam's defense, shipbuilding, and aerospace sectors will translate into production and workforce demand between 2026 and 2028. If current talent education aligns with this demand, graduates will enter the industrial boom; however, if it does not, companies will fill the gap with external experienced workers, foreign personnel, or automation. There is currently no evidence linking order schedules with the majors, competencies, and hiring timing of graduates from specific schools. Entry pathways for young people that were not established during a boom period narrow further during a recession.  The current industrial boom is the final favorable market segment to assess the suitability of Gyeongnam's education for Manufacturing AX.

12-1. Golden Time Application Case in Basic Local Governments ① — Yangsan City

Yangsan AI High School opened in 2025 with four departments—AI Convergence Factory, AI Automatic Control System, AI Content, and Bio-Food—and a total of 378 students across 21 classes, with 126 students per grade. The first recruitment attracted 302 applicants, recording a competition ratio of 2.4 to 1, and practical training facilities simulating actual industrial sites were also established. While student demand for specialized AI education and physical infrastructure have been confirmed, actual data from manufacturing companies in Yangsan and Gimhae, joint projects, employment agreements, and post-graduation retention rates have not yet reached the verification stage. If this gap persists until the first graduating class is produced in 2027–2028, the school’s AI brand will be disconnected from the demand for AX personnel in the local manufacturing industry.  Yangsan’s "Golden Time" is not the opening of the school, but the period to determine whether the first graduating cohort is absorbed by the local manufacturing sector. ( Gyeongnam Office of Education )

12-2. Golden Time Application Cases in Basic Local Governments ② — Changwon City

Changwon Technical High School for Mechanical Engineering was selected for the 2026 Agreement-Type Specialized High School and Department Restructuring Projects, and plans are underway to expand practical training tailored to local industries, centering on the AI ​​Smart Machinery Department. With 3,216 companies and over 120,000 employees in industrial complexes, Changwon boasts the largest physical foundation for corporate cooperation in Gyeongnam. However, evidence regarding the percentage of joint curriculum incorporating tasks required by prime contractors and subcontractors in the defense, nuclear, and machinery sectors, as well as student project participation, conversion to employment, and three-year tenure, has not yet been presented. If the Agreement-Type Specialized High Schools operate from 2026 to 2028 based on company lists and facility investments, Changwon's industrial scale will not translate into educational outcomes.  Changwon is not a region lacking companies, but rather a region undergoing verification to determine whether it has transformed corporate concentration into a career ladder for its students. ( Yonhap News )

The employment rate for vocational high schools in Gyeongnam Province is projected to reach 55.9% in 2025, slightly higher than the national average but a 3.6 percentage point drop from the previous year. At the same time, the manufacturing sector is being reorganized around AI, automation, and digital quality. If the gap between education and the labor market persists, graduates will fail to secure their first jobs, while companies will repeatedly face a shortage of experienced workers. Failure to find an initial job accumulates into disparities in job experience, wages, and regional retention. A delay of two to three years in entering the workforce will translate into a shortage of mid-level skilled workers around 2030.  What is currently being overlooked is not the number of employed individuals, but the next skilled generation of the Manufacturing AX.

In industries highly exposed to AI, youth employment has decreased while employment for those in their 50s has increased. If this bias toward experienced workers extends to the manufacturing sector, companies will shift toward integrating AI with existing skilled workers and reducing the hiring of new youth. Schools lack evidence distinguishing between entry-level tasks that have disappeared due to automation and those that have newly emerged. After entry-level roles are reduced between 2026 and 2028, students will not be able to generate practical field experience at all. What  Gyeongnam education is missing is not AI knowledge, but the first opportunity for young people to grow into skilled professionals.

If specialized manufacturing education is excessively tailored to a single anchor company or current orders, skill mobility decreases during industrial downturns. Although Gyeongnam possesses multiple industries—machinery, defense, shipbuilding, and aerospace—there is a lack of common competencies among students and evidence for job transfer between industries. If closed-system education by industry becomes entrenched for two to three years, a shock to one industry directly leads to graduates migrating out of the province.  The inter-industry mobility that is not currently established reduces the likelihood of recouping educational investments in the face of future economic shocks.

iTalkTalk’s learning data exceeding 7TB, 89% of schools registered for AI Math, 18 restructured departments across 14 schools, Yangsan AI High School, the Vocational Education Innovation District, and RISE coexist. Basic points of contact for determining the connection between education and industry have already been established. However, there is no evidence that spans the learning, projects, employment, tenure, and regional retention of the same students. Once the timeline of the initial cohort is secured between 2026 and 2028, a baseline for comparing distributed projects will be created for the first time.  The greatest asset currently available is not new educational facilities, but the observability of the Manufacturing AX talent pathway.

Changwon, Yangsan, Gimhae, Geoje, and Sacheon possess distinct manufacturing industries and corporate structures, and new academic departments, such as AI Energy, have emerged in the western inland region. Regional conditions exist that allow the same foundational AI competencies to transfer to smart machinery, aviation, shipbuilding, and energy sectors. However, actual inter-industry mobility of students and the recognition of skills between companies have not yet been confirmed. If evidence of this mobility is generated within two to three years, industry-specific education will be recognized as a provincial career market rather than a closed major.  The structural asset currently available is not a workforce tailored to a single industry, but rather skills whose value is maintained across multiple sectors.

The first graduation outcomes of the reorganized departments, including Yangsan AI High School, will appear in 2027–2028. The initial cohort is the group that allows for the clearest comparison of the effects of curriculum, faculty, equipment, and corporate collaboration. If data on job placement, wages, tenure, and regional retention from this period are omitted, subsequent cohorts will be mixed with additional policies and economic fluctuations, making causal determination difficult.  The time assets currently available serve as a baseline for verifying the performance of the first graduating class without distortion.

The scope of AX response is not limited to the addition of AI subjects and coding classes.  It is a structure that connects the curriculum → industrial tasks → projects → first employment → tenure → job transition → regional retention into a single Evidence Chain. Even if the AI ​​utilization rate in schools increases, if employment in related jobs and retention of employment do not improve, it is not evaluated as manufacturing AX talent performance.

The first Runtime sets students from Yangsan AI High School and the recently restructured 18 departments as the baseline cohort. Competencies at the time of admission, projects by grade level, corporate evaluations, job roles, starting salary, 1-year and 3-year tenure, and retention in Gyeongnam are connected along the same time axis. For general high school students, the path connecting iTalkTalk's math, science, and information learning to the Manufacturing AX career path is tracked separately.

The cause of failure is not a lack of AI classes. Although department names and equipment have changed, the on-site adaptation period at companies has not decreased, the employment and retention rates of vocational high schools in Gyeongnam remain stagnant, and graduates of the reorganized departments are moving to other majors and regions instead of manufacturing AX roles.

Educational Infrastructure Opportunity + Industrial Talent Pipeline Risk

Gyeongnam Education has secured a leading position in the stage of disseminating AI to schools. However, results have not yet been proven in the stage of supplying talent capable of solving manufacturing problems using AI to the industry.

The current danger is not the absence of AI education. It lies in the fact that while curriculum platforms, vocational high schools, universities, and companies operate independently, the pathway leading students' competencies to their first job and long-term careers remains empty.

The period from 2026 to 2028 is the Golden Time when the first results of the reorganized departments and AI-specialized schools will be determined. Once this period passes, it becomes difficult to separate the failure of the curriculum from the impact of the economic downturn and the outflow of youth.

Final judgment: AI learning infrastructure is high, diffusion of future industry departments is medium-high, job connection in manufacturing is medium-low, tracking of long-term performance is low, and time risk is high.

The verification order is Learning → Practical Training → Corporate Project → First Employment → 1 Year Retention → 3 Years Job Growth → 5 Years Remaining in Gyeongnam . If any one of the number of registered schools, reorganized departments, or partner companies increases, it will not be judged as Manufacturing AX talent performance.

Runtime Chain

Student Learning Evidence → Manufacturing Project → Company Verification → Entry Job → Skill Growth → Regional Retention

Structural insights remaining from this analysis

The structural gap in Gyeongnam education does not lie in a lack of AI education. Gyeongnam already possesses a wide-area AI learning platform, over 7TB of learning data, 89% of schools registered for AI mathematics, and AI-specialized high schools and departments for future industries. What is lacking is connected evidence proving which problems in the manufacturing industry students who have learned AI have solved .

AI learning platforms record what students know, vocational high schools record what equipment students handled, and companies record what problems occurred on the production floor. While these three records remain separate, education creates achievement levels, and companies then provide on-the-job adaptation training. Although educational investment and corporate training costs overlap, there is no record of at which stage a gap occurred.

In the era of Manufacturing AX, a history of problem-solving serves as a stronger hiring signal than the name of the academic department. Proficiency is demonstrated not by whether one graduated from an AI Smart Machinery department, but by whether they predicted failures using actual equipment data, classified defects through video inspection, or increased productivity by adjusting process variables. Without this transition, the spread of AI education and the labor shortage in the manufacturing sector will continue to coexist without conflict.

Golden Time Thesis — The Golden Time for Gyeongnam education from 2026 to 2028 is not a period of teaching more AI, but a period of transforming students' learning data into actual problem-solving history for Gyeongnam's manufacturing industry. Once this connection is formed, iTalkTalk, vocational high schools, RISE, and companies will become a single talent supply chain. If this connection is not established, Gyeongnam will simultaneously face the contradiction of being a leading region in AI education while suffering from a shortage of manufacturing AX talent.