Busan’s AI Bet Will Be Judged by What Stays

The Yeongnam advanced-industry blueprint puts Busan on Korea’s AI map. The harder test is whether port AI, manufacturing AX and Samsung Electro-Mechanics’ 15 trillion won investment leave behind local skills, operating jobs, supplier capacity and firms that can keep the systems running.

Busan does not need to be persuaded that AI is coming for its port, factories and maritime economy; that argument is already over. A major logistics city cannot ignore automated terminals, predictive maintenance, AI safety monitoring, manufacturing data systems, power semiconductors or server components and still expect to remain competitive. The harder question, sharpened by the National Briefing on the Vision for Advanced Industry Development in the Yeongnam Region in Jinju on July 3, is whether Busan will own enough of that transition to matter.

The government and major conglomerates presented the briefing as part of a 312 trillion won regional investment package covering AI data centers, next-generation semiconductors and materials, physical AI, robotics, aerospace, defence, energy and advanced manufacturing. For Busan, the most concrete new item was Samsung Electro-Mechanics’ plan to invest about 15 trillion won in its Busan site by 2040 to strengthen AI data-center server package substrates and multilayer ceramic capacitors, with the company describing the city as a core production and R&D base for high-performance package substrates and MLCCs.

That is a serious development because it gives Busan a clearer place in the national AI supply chain, not only as a port city talking about smart logistics, but as a production and research base for components that will sit inside the infrastructure of AI itself. It also connects the city’s older manufacturing base in the western industrial belt to a newer language of data centers, advanced components and intelligent production. Yet a place on the map is not the same as control over the value chain. A city can host AI systems without becoming an AI economy. It can install smarter cranes, fund factory demonstrations, attract component investment and announce global ambitions while the software, maintenance contracts, operating knowledge, data rights and higher-value jobs remain somewhere else. In that case, Busan would not be transformed by AI so much as used as a site for systems designed, governed, upgraded or monetized outside the region.

That is the real test behind the city’s maritime-capital agenda. Busan is right to build its future around the port, maritime industries, manufacturing AX, AI infrastructure components and data systems. The port already gives the city a central place in Korea’s physical trade economy. Containers move through its terminals, logistics firms depend on its schedules, ship-equipment suppliers cluster around its western industrial belt, and national growth plans continue to describe Busan as a southern gateway. The weakness has never been a lack of activity. It has been the location of the higher-value functions around that activity: contracts, finance, insurance, legal work, data systems, maintenance knowledge and technical decisions that often sit outside the city even when the cargo, factories and workers are in Busan.

AI will either correct that pattern or deepen it. If software is controlled elsewhere, maintenance contracts are written elsewhere, public data becomes a testing ground for outside vendors, component investment creates production without a wider supplier and skills ecosystem, and young workers receive training without access to entry-level technical jobs, Busan will have hosted the transition without capturing much of its value. The city does not need to slow its AI transition. It needs to make the transition harder to extract from Busan, and the measure should be what remains after the system is installed or the investment is announced: local firms, local maintenance teams, local data capacity, local supplier growth, local first-rung technical jobs and institutions capable of governing the maritime and manufacturing economy rather than merely hosting it.

A Regional Number Is Not a Local Strategy

The Yeongnam advanced-industry blueprint gives Busan scale, but not yet capture. That distinction matters because regional investment announcements can make every city inside the map look included. A 312 trillion won package covering Yeongnam can make Busan appear part of Korea’s next industrial frontier, but the actual geography of value will be decided by what each city receives, which functions are placed there, which firms grow around them and whether local workers can enter the new occupations that follow.

The government’s regional blueprint divides Yeongnam across several industrial roles. SK’s 140 trillion won plan includes a 2GW-class AI data-center push in the region, with Ulsan named for a 1GW mega data center. Samsung’s 60 trillion won package includes humanoid robot production, advanced batteries and AI manufacturing. Hyundai Motor Group’s 42 trillion won commitment points toward AI-based autonomous mobility, future parts clusters, manufacturing AI and aerospace mobility. Hanwha’s 55 trillion won package emphasizes satellites, launch vehicles and space-defence AI infrastructure. The government also described a Busan power-semiconductor cluster, a Gumi semiconductor materials and defence-oriented testbed, a Gumi-Pohang-Daegu-Changwon advanced robotics belt, and a broader effort to bring physical AI into manufacturing, logistics and defence.

In that map, Busan’s most concrete new claim is not that it will host the largest AI data center or the main humanoid robot production line. It is that Samsung Electro-Mechanics’ Busan site will be strengthened as a high-performance package substrate and MLCC base for AI data-center servers, while the government continues to present the city as a power-semiconductor and maritime-industrial node. That is not a small role. AI infrastructure needs far more than algorithms and cloud halls. It needs substrates, capacitors, power devices, precision manufacturing, quality control and supplier discipline.

The opportunity also defines the risk. Busan could become a more important production node inside a national AI infrastructure strategy while still missing the higher-value layers attached to that role. The city’s question is not only whether 15 trillion won is invested, but whether the investment anchors R&D work, process engineering, testing capacity, supplier upgrading, training programs, quality-control jobs, maintenance knowledge and smaller firms able to earn recurring business around the plant. A regional number is useful for national industrial politics, but it is not a local economic strategy. Busan’s task is to translate its place in the package into functions that remain in the city, which requires a different vocabulary from the usual language of investment attraction. The better question is not how much has been announced for Yeongnam, but what part of the value chain Busan can govern, maintain, repair, improve and teach.

A Maritime Capital Is a Function

The phrase “maritime capital” can sound like a status claim, as if the city is asking for recognition of what geography has already given it. Busan does not need to prove that it is connected to the sea. That fact is visible in its port, ship-equipment suppliers, logistics firms and industrial districts built around maritime trade. The more difficult step is to turn that physical base into a higher-value institutional economy.

A maritime capital is not simply the city where cargo passes through. It is the city where maritime decisions are made, marine risks are priced, shipping disputes are settled, logistics data is governed, systems are maintained and workers are trained into the industries built around the port. A port city that only handles cargo remains exposed to shipping cycles, automation and decisions made elsewhere. A maritime capital should be harder to bypass.

That is the strongest reading of Mayor Jeon Jae-soo’s maritime agenda. The relocation of oceans administration, the proposed maritime court, shipping-company clustering, maritime finance and AI-enabled port operations should not be viewed as separate trophies. Their economic value depends on whether they reinforce one another. A ministry can bring administrative gravity. A maritime court can build legal specialization. Shipping companies can generate headquarters demand. Finance and insurance can price the risks behind vessels, cargo, terminals and the energy transition. AI and port data can make Busan not only a place where logistics happens, but a place where logistics systems are designed, audited and improved.

The Yeongnam advanced-industry blueprint makes this function more important, not less. If the wider region is being reimagined as an advanced manufacturing and physical-AI belt, Busan’s maritime role cannot remain a ceremonial identity. It has to become an operating platform linking ports, manufacturing, components, ship equipment, logistics software, power devices, data governance and technical labour.

This is where the opportunity is real. Busan already has the industrial surface on which such a strategy can work. Its port is not an imagined asset. Its western industrial belt is not a blank development zone. Its maritime identity is not a campaign slogan invented from nothing. The city’s problem has been that too much of the value attached to those assets has been captured elsewhere. A container may move through Busan, but the financing, legal work, software architecture, insurance decision, platform contract and equipment-maintenance margin can sit outside the city. A supplier in Noksan may make parts for a national industry, while the digital system that raises productivity is sold, updated and controlled by a vendor with little reason to leave technical capacity in Busan.

The Global Hub City framework shows why the distinction matters. The special-law push has been framed around international logistics, finance, advanced industry, special zones, tax incentives and regulatory exemptions intended to position Busan as a southern global hub. Those provisions may open a door, but they do not create a city economy by themselves. Busan’s gain will depend on whether special zones and institutional relocation produce a market around maritime law, insurance, finance, data services, logistics platforms and technical operations, rather than another layer of policy status over an unchanged value chain.

The same logic applies to the city’s new AI language. Busan is unlikely to become Korea’s main semiconductor production region, and it does not need to pretend otherwise. Its stronger claim is as an application and operating city, where AI meets ships, terminals, factories, warehouses, cold chains, customs flows, safety systems, power devices and manufacturing processes. That position is less glamorous than chip fabrication, but it is more relevant to Busan’s own economy. The city can become valuable if it turns AI from a national export story into local operating capacity.

West Busan Has Heard This Promise Before

West Busan is where Busan’s AI language will face its most unforgiving test. The machinery, ship-equipment and parts manufacturers clustered around Myongji and Noksan are not waiting for a slogan. They are dealing with older factories, thin margins, labour shortages, delivery pressure, quality requirements from larger customers and the steady demand to produce more reliable parts with fewer errors.

AI can matter in that environment. A vision-inspection system that lowers defects, a digital work instruction that reduces rework, a predictive-maintenance tool that keeps a machine from stopping during a delivery cycle, or a production-data system that helps a small supplier prove quality to a buyer can be the difference between remaining in a supply chain and being pushed out of it. That is why manufacturing AX should not be dismissed. West Busan needs industrial upgrading, and the Myongji-Noksan belt is too important to be treated as yesterday’s economy. City and project materials have described the area as the country’s largest ship-equipment cluster and a core industrial base accounting for 28.5 percent of Busan’s manufacturing production and 32.2 percent of its exports. The zone is not a peripheral workshop, but one of the places where Busan’s manufacturing base is still concentrated.

The difficulty is that West Busan has already heard a version of this promise. Noksan was placed inside the smart-green industrial-complex agenda in 2022, when Busan launched a five-year program through 2026 with 15 projects covering digital transition, smart-factory advancement, energy self-sufficiency, smart convenience facilities and industrial-environment improvement. The language was different then, but the ambition was familiar: make an ageing industrial base cleaner, smarter, more productive and more attractive to workers.

The new AX phase therefore cannot be written as a clean break from the past. The Myongji-Noksan AX demonstration project selected in 2025 will invest 24.9 billion won through 2028 and aims to support technology testing and diffusion among small and medium-sized manufacturers. Its focus on digital twins, eco-marine materials, parts and equipment, and technology diffusion is sensible, but it also shows how basic the work still is. Before AI can transform production, many firms still need diagnosis, standardization, data preparation and training.

That history should make the city more exacting, not more cynical. A small manufacturer does not become AI-ready because a district receives a new policy label. It needs machine data that can be trusted, inspection records that can be used, sensors that are maintained, workers who understand the system, vendors that do not disappear after a pilot, and enough cash flow to keep the tool running after the subsidy ends. The older language of smart factories often assumed that digital tools would spread once infrastructure and support programs were made available. The harder lesson is that adoption inside small firms is slow because it touches production routines, worker training, maintenance budgets, customer requirements and management capacity at the same time. A pilot is not adoption unless the factory can keep using it on an ordinary Tuesday, with its own workers, its own production data and a maintenance structure it can afford.

West Busan’s AX project should begin from that lesson. The measure should not be how many firms receive a diagnosis, how many demonstration systems are installed or how sophisticated the pilot looks during a launch event. The measure should be whether firms keep using the systems after one year, whether defect rates fall, whether delivery reliability improves, whether energy and downtime costs decline, whether local suppliers gain recurring technical work, and whether younger workers find entry-level roles as maintenance assistants, production-data coordinators, quality-system operators or manufacturing-AI technicians.

Samsung Electro-Mechanics’ 15 trillion won Busan investment raises the stakes around this question. If the investment remains a large corporate production and R&D project isolated from the wider local manufacturing base, its impact will be real but contained. If it helps upgrade local suppliers, technical schools, engineering pathways, process-control skills and component-related service firms, it can become part of the city’s industrial thickening. That is the difference between a plant and an ecosystem. West Busan does not need another demonstration that technology can work in principle. It needs operating evidence that technology can remain inside ordinary manufacturers after consultants, grants and announcements have moved on.

Port AI Needs a Labour Test

Busan Port cannot stand outside automation. A container port that ignores AI scheduling, automated terminals, predictive maintenance, truck-flow management and safety analytics will not protect its workers for long; it will slowly weaken the industrial base that supports them. The largest ports are no longer competing only on berth depth, crane capacity or location. They are competing on how quickly they can predict congestion, recover from disruption, coordinate trucks and vessels, reduce accidents and move data across terminal operators, logistics firms and public agencies. For Busan, port AI is not a decorative technology agenda. It is part of remaining relevant as a logistics city.

That is why the Busan Port AX plan deserves to be treated as a serious industrial project. BPA has described a plan with 38 detailed tasks through 2030, a total project scale of 892.1 billion won, including 435.1 billion won from its own budget, with the goals of raising container-terminal productivity by 30 percent and pushing port safety toward zero fatal accidents. A port where trucks wait less, cranes fail less often, vessels are scheduled more accurately and dangerous work is monitored more intelligently is a better port. For a city whose economy still depends on maritime logistics, those gains matter.

The employment case is more complicated. A more productive port is not automatically a larger labour market. AI can reduce waiting time without increasing headcount. Automated handling can shift work away from yards and gates into control rooms. Predictive maintenance can prevent breakdowns while changing the skills required of maintenance teams. AI safety systems can reduce accidents while also reducing some forms of manual monitoring. None of this means port AI should be rejected. It means the city should stop treating productivity gains as if they already answer the jobs question.

Busan’s port transition will create work only under certain conditions. The new work is likely to sit in remote equipment operation, system maintenance, AI safety monitoring, data-quality management, cybersecurity, logistics coordination and equipment diagnostics. Those are real jobs, but they are not the same jobs that automation may reduce. They require different training, different wage structures, different certification and different relationships between terminal operators, vendors and public agencies. If the new roles are written for experienced engineers from outside the region, or handled by vendors whose teams are based elsewhere, Busan will see smarter terminals without necessarily seeing a stronger local labour market. A smarter terminal is not a stronger local economy unless the technical work around it is also rooted locally.

The city has one advantage: port automation is visible enough to be negotiated. A factory may introduce software quietly inside one firm, but a terminal transformation affects workers, operators, transport firms, safety systems and the public port authority at the same time. That gives Busan a chance to make labour transition part of the operating design rather than an afterthought. The city and BPA can ask, before each system is installed, which existing tasks will shrink, which tasks will move, which new roles will be created, where those roles will be based, and whether current workers can be trained into them.

That is the difference between buying automation and building capability. If an AI terminal system arrives as a vendor package, Busan may receive the hardware, the dashboard and the performance claim while losing control over the knowledge that makes the system useful. Software updates, maintenance protocols, data standards, cybersecurity decisions and equipment diagnostics can all remain outside the city. The port becomes more advanced, but the city remains dependent. If the contract requires local maintenance teams, technical documentation, worker retraining, data-access rules and university-linked practice, the same technology can become part of Busan’s labour-market infrastructure.

Busan’s maritime-capital strategy depends on this distinction. If the port becomes more intelligent while the intelligence is owned, maintained and upgraded elsewhere, Busan remains a place where trade passes through. If the port becomes more intelligent and the city also builds the workers, firms and institutions that operate that intelligence, Busan moves closer to becoming the maritime decision-making city it says it wants to be.

Impact Numbers Need a Labour-Market Audit

The Yeongnam advanced-industry package gives Busan a larger number to stand under, but it does not yet give Busan a local jobs account. That distinction is not a criticism of the package itself. Large investment announcements have a role in industrial policy. They show political commitment, corporate alignment and the scale of infrastructure that may be required for AI, robotics, advanced components, aerospace and energy. They can help justify public support, regulatory changes, tax incentives and new financial vehicles. A city competing for national attention needs to show more than aspiration. The problem begins when an impact figure is treated as if it already answers the labour-market question.

A 312 trillion won regional package does not tell a young worker in Busan what job will exist in the city five years from now. A data-center number does not explain whether the operating work will be in Busan, Ulsan or somewhere else. A component investment does not automatically show whether local subcontractors will grow around it. A manufacturing AX budget does not reveal whether small firms will keep systems running after grants expire. A regional employment projection does not tell whether jobs are construction-phase, indirect, induced, temporary, senior specialist or first-rung local positions.

This distinction also applies to the proposed UN AI hub. The Busan Research Institute has presented the city’s submarine-cable connectivity, energy self-sufficiency, port, digital-asset infrastructure and Eco Delta City as location assets, and has projected large production and employment effects across the southeast if such a hub is located in Busan. Those are useful bid numbers. They are not yet a local jobs account. Induced output is not local income, and cumulative employment effects do not automatically mean permanent jobs in Busan. A forecast for the wider southeast region may include construction, services, events, indirect supplier activity and spending effects spread across several places, while saying much less about durable operating roles inside Busan itself.

A young worker deciding whether to remain in the city is not helped by a headline employment number unless that number can be translated into actual entry points: AI-governance assistants, data-operations staff, cybersecurity technicians, maritime-logistics analysts, compliance workers, cloud-infrastructure operators, component-process technicians or public-sector AI specialists with employers, wages and career paths attached.

That is why these estimates should be audited rather than dismissed. A serious jobs case would separate direct employment from indirect and induced employment, construction-phase activity from permanent operation, Busan’s share from the wider region, and local-firm contracts from spending captured by national or global vendors. It would also distinguish between senior specialist roles recruited from outside the region and first-rung positions that local graduates or retrained workers could enter. Such an audit would not weaken the bid. It would make it more credible, because a city that can explain the structure of its employment impact is making a stronger case than one relying on the size of the headline number alone.

The same standard should govern the rest of Busan’s AI agenda. Busan Port AX should be measured not only by terminal productivity and safety targets, but by the operating roles it creates in remote control, maintenance, data management, equipment diagnostics and cybersecurity. West Busan’s manufacturing AX should be measured not only by the number of firms diagnosed or pilots launched, but by installation rates, one-year system retention, internal operating staff, defect reduction and delivery reliability. Samsung Electro-Mechanics’ Busan investment should be measured not only by capital expenditure, but by its effect on local R&D, suppliers, training, engineering work and durable technical jobs. A youth AI hub should be measured less by training capacity than by the first jobs its graduates can actually enter. Large projects can justify ambition, but they cannot by themselves show who works, where the work sits, which firms are paid, what skills are transferred and how much value remains in Busan after the announcement cycle ends.

The Missing Metric Is Local Capture

Busan’s most important metric is not the size of the project. It is the share of value the city keeps after the project is announced, installed and normalized. Local capture is the difference between hosting a project and building an economy. A ministry can relocate without moving the private market around it. A headquarters can change its address while procurement, legal advice, finance and strategic decisions remain elsewhere. A smart-port system can be installed in Busan while the software, maintenance protocols, data architecture and upgrade contracts are controlled from outside the region. A manufacturing AX pilot can use local factories as demonstration sites while recurring revenue, engineering knowledge and client relationships flow back to vendors with little permanent presence in the industrial belt. A component investment can raise production volume while the higher-value decisions around design, equipment, materials, testing and supplier development remain thin locally.

Busan’s maritime-capital strategy will not be proven by the number of institutions it attracts or the number of AI systems it deploys. It will be proven by whether those institutions and systems create a thicker local market for work that did not previously sit in the city. Maritime finance, marine insurance, shipping arbitration, port-data governance, smart-terminal maintenance, manufacturing-AI support, cybersecurity, logistics software, equipment diagnostics, component testing and compliance services are not symbolic add-ons to the port economy. They are the higher-value functions that determine whether Busan remains a place where trade passes through or becomes a place where maritime value is organized.

That distinction should make the city more selective about what it counts as success. A port AI platform that reduces congestion is valuable, but it becomes more valuable to Busan if local engineers maintain it, local firms support it, local universities train workers on it and local data-governance rules determine how it is used. A manufacturing AX program that lowers defects inside a supplier is valuable, but it becomes more valuable to Busan if the supplier can keep the system running without permanent dependence on outside consultants. A large AI-server component investment is valuable, but it becomes more valuable to Busan if it strengthens local process knowledge, supplier capability, testing services and technical career paths. A youth AI hub is useful only if it connects training to identifiable first jobs in the port, factories, logistics firms, public agencies, component plants and AI-service companies that actually operate in the city.

The danger is not that outside firms participate. Busan will need national technology companies, global cloud providers, port-equipment makers, component leaders, software vendors and specialized consultants if it wants to modernize quickly. External expertise can accelerate an industrial transition. The danger is a contract structure in which outside firms take the durable value while Busan supplies the site, the data, the workers and the political narrative. Outside participation becomes extraction when operating knowledge, data rights, maintenance work and supplier margins never take root locally.

A serious local-capture strategy would change the questions asked at the beginning of each project. Who will own or govern the data after deployment? Which parts of maintenance must be performed in Busan? What documentation will be transferred to public agencies, local firms or participating manufacturers? Which subcontracting opportunities are reserved for local suppliers capable of growing beyond the pilot? What entry-level roles will be created for workers who do not already have years of AI, robotics or semiconductor experience? Which university labs and vocational programs will work with live industrial problems rather than generic training modules? Which component suppliers can move from low-margin production into testing, quality, materials or process-support work? These are not administrative details, but the mechanisms through which an AI project becomes an economic base.

The test is whether Busan can make itself difficult to bypass. A city is bypassed when cargo moves through it but decisions are made elsewhere, when data is generated locally but monetized elsewhere, when workers are trained locally but hired elsewhere, when factories host experiments without gaining the capacity to operate the next generation of production systems themselves, and when large plants operate in the city without thickening the surrounding supplier and skills ecosystem. A maritime capital worthy of the name would reverse that pattern. It would make Busan the place where maritime problems are defined, systems are maintained, contracts are written, risks are priced, components are tested, workers are trained and firms earn recurring revenue from the industries built around the port.

The Contract Is the Strategy

The decisive document in Busan’s AI transition will not be a vision statement. It will be the contract. Maritime-capital language can set direction, and AI strategies can identify promising sectors, but the distribution of value will be decided in procurement clauses, investment agreements and operating arrangements that determine who controls data, who maintains equipment, who updates software, who trains workers, who receives recurring service revenue and who can keep the system running after the first deployment.

A city that buys AI as a finished product gets dashboards, hardware and performance claims. A city that buys AI as operating capacity gets people, documentation, maintenance rights, local suppliers and institutional knowledge. Busan’s port and manufacturing projects should be designed with that difference in mind from the beginning. A smart-terminal system can improve scheduling and safety while leaving the city dependent on an external vendor for every upgrade, fault diagnosis and security patch. A factory-inspection tool can reduce defects during a pilot while leaving the manufacturer unable to retrain the model, repair the sensors or integrate the data with the rest of its production process. A public AI platform can be described as shared infrastructure while smaller firms remain unable to afford access, understand pricing, trust data governance or hire workers capable of using it. In each case, the technology may function, but the local economy remains thin because the operating capability has not been transferred.

Procurement can correct that weakness only if Busan treats it as part of industrial policy. Every major port AI, manufacturing AX or maritime-data contract should define more than deliverables. It should define the local operating structure around those deliverables: which maintenance tasks must be performed in Busan, which technical documents must be transferred, which data standards must be open to participating firms, which cybersecurity responsibilities stay with public agencies or local operators, which local suppliers are trained into the system, and which workers receive credentials that remain useful beyond a single vendor’s platform. Without those conditions, public money can modernize assets while failing to modernize the city’s own capacity.

West Busan’s manufacturers make this especially urgent. Smaller firms do not need a one-time demonstration as much as they need a maintainable operating system. A supplier in Noksan that receives an AI inspection tool still has to collect clean production data, keep cameras and sensors calibrated, train workers to interpret false positives, connect quality records to customer requirements and pay for support after the subsidy period ends. A public program that funds only the first installation may produce a visible success story and still leave the firm unable to continue. A better contract would ask what happens six months later, a year later and after the vendor’s initial support expires.

The same logic applies at the port. Automated cranes, AI safety monitoring and predictive maintenance can make Busan Port more competitive, but the city’s labour-market gain will depend on whether workers move into the technical layers of that system. Remote operation, equipment diagnostics, safety analytics, data-quality management and cybersecurity should not be treated as incidental byproducts. They should be written into the transition as occupations with training pathways, wage ladders and local hiring expectations. If the new work is treated as a specialized service purchased from elsewhere, Busan will have improved the terminal while weakening the link between the terminal and the local labour market.

Investment agreements around advanced components should face a similar test. Samsung Electro-Mechanics’ Busan investment can be a major anchor for the city’s AI-infrastructure role, but the broader local effect will depend on whether the investment creates supplier openings, technical training, process-engineering pathways, R&D collaboration and quality-control work that remain in Busan. A plant can be world-class and still be locally thin. The point is not to burden investment with vague social expectations, but to ask how a city can convert a major corporate commitment into a broader industrial base.

This is why procurement and investment governance should be understood as the practical core of the maritime AI strategy. Speeches can say Busan will become a maritime capital. Contracts decide whether the city can operate like one. A maritime capital cannot depend indefinitely on outside actors to maintain its port intelligence, interpret its logistics data, update its industrial AI systems, support its component base or train its technical workforce. It needs the ability to govern, operate and repair the systems that will increasingly define maritime and manufacturing competition. The contract is not a bureaucratic detail; it is where Busan decides whether AI remains a visiting technology or becomes part of the city’s economic machinery.

Busan Should Publish the Jobs Math

Busan’s AI transition will remain vulnerable to suspicion until the city makes its employment logic visible. A public that has heard years of smart-city, smart-port and smart-industrial-park language will not be persuaded by another promise that technology will eventually create better jobs. Workers in the port, managers in West Busan factories, young graduates looking at Seoul, and small firms asked to adopt unfamiliar systems all need a more concrete account. They need to know which work will disappear, which work will change, which work will be created, and which of those jobs will actually be based in Busan.

The city does not need an impossibly precise forecast. Industrial transitions are uncertain, and no administration can know exactly how many technicians, data operators, safety monitors, process engineers or logistics-software workers will exist five years from now. But uncertainty is not a reason to avoid accounting. It is a reason to publish assumptions. A port AI project should state whether productivity gains are expected to come from faster handling with the same workforce, from workforce reduction, from redeployment into control-room and maintenance roles, or from increased cargo volume that supports additional work. A manufacturing AX project should say whether public support is designed to create new jobs, defend existing jobs by keeping firms competitive, or move current workers into more technical roles. A component investment strategy should identify which roles are production, which are R&D, which are process engineering, which are quality control, which are supplier support and which can become first-rung jobs for local graduates. A youth AI hub should identify not only the number of people trained, but the employers, job titles and wage levels its graduates are expected to enter.

That distinction matters because “job creation” can mean too many things at once. A construction worker hired during the build-out of an AI facility is not the same as a permanent systems operator. A consultant brought in for a pilot is not the same as a local technician employed by a manufacturer. A worker retained because a factory becomes more competitive is not the same as a new graduate entering an AI-related job. A senior engineer recruited from outside the region is not the same as a first-rung local role that gives a young worker a reason to stay. Each of these outcomes has value, but they should not be collapsed into a single employment figure. Busan’s problem is not that every AI project must produce large new headcounts. The problem is that the city cannot judge its own strategy if it does not separate job creation, job defence and job transformation.

A credible jobs account would begin with occupations rather than slogans. For Busan Port AX, the city and port authority should map how automation affects yard work, gate operations, crane operation, safety monitoring, equipment maintenance, data management and cybersecurity. For West Busan manufacturing AX, they should map how AI affects inspection, quality control, machine maintenance, production planning, energy management, documentation and customer compliance. For maritime AI and global-hub projects, they should map roles in data governance, maritime standards, AI safety, public-sector AI, logistics analytics, marine insurance, ship finance and regulatory compliance. For AI-server components and power semiconductors, they should map roles in process control, materials testing, substrate production, MLCC quality, equipment maintenance and supplier engineering. This would not turn an op-ed concern into a bureaucratic exercise. It would force each project to say what kind of labour market it is actually building.

Young workers should be treated as a separate line in that account, not as a rhetorical afterthought. Busan’s demographic problem makes this unavoidable. A city that loses young people cannot call an AI strategy successful only because it trains workers already employed elsewhere or attracts senior specialists for a limited project. The strategy needs first-rung jobs: assistant roles in port data operations, junior equipment-diagnostics positions, manufacturing-data coordinators, AI safety-monitoring staff, smart-logistics schedulers, cybersecurity support roles, maritime-compliance analysts, component-quality technicians and maintenance workers who can learn inside the system. Those jobs may not sound as glamorous as an AI hub or a 312 trillion won regional package, but they are the positions that turn training into a labour market. Training is not a labour market unless there is a first job at the end of it.

Publishing the jobs math would also improve political honesty. It would allow Busan to say that some AI projects are primarily productivity projects, some are job-defence projects, some are workforce-transition projects, and only some are direct job-creation projects. That distinction would make the city’s case stronger, not weaker. Residents can accept automation more easily when the trade-offs are visible and the pathway for workers is real. Firms can trust public programs more when they know support extends beyond diagnosis and launch events. Young people can judge whether an AI strategy offers a career or only another training certificate.

A city that wants to become a maritime capital cannot treat work as an afterthought to technology. The port, the industrial belt, the component plants and the institutions around them will be judged by whether they create a more durable local economy, not by whether they sound advanced on paper. Busan should not ask citizens to believe that AI will create jobs in the abstract. It should show, project by project, where the work will be, who can enter it, who will maintain it, and how much of it will remain in the city.

The Work Must Remain

The Jeon administration’s opportunity is that Busan’s next industrial story does not have to be invented from scratch. The city has a port that still matters, a western manufacturing belt that still produces, a maritime identity that is more than decorative, and now a clearer link to the AI infrastructure supply chain through major component investment. It also has a policy opening around maritime administration, global logistics, AI infrastructure, regional finance and advanced manufacturing. Those assets give Busan a more serious foundation than the usual language of urban reinvention, but they also make the standard higher. A city with real industrial weight cannot be satisfied with projects that sound modern but leave the structure of value unchanged.

The first obligation is continuity with honesty. West Busan’s manufacturing AX should not pretend that smart-industrial-park policy began yesterday. Earlier rounds of smart factories, smart-green industrial complexes, regeneration projects and manufacturing innovation programs created a record that should now be examined rather than buried under new terminology. The city should say which investments worked, which firms upgraded, which systems remained in use, which bottlenecks persisted and why the new AX phase will be different. A policy that admits what the previous round did not solve will be stronger than one that simply changes the name of the promise.

The second obligation is to make port automation a labour transition before it becomes a labour conflict. Busan Port will have to automate, and the city should say that plainly. The harder task is to make the transition visible to the people whose work will be reorganized by it. Terminal productivity, safety analytics and autonomous handling should be accompanied by a map of occupations: what happens to yard work, gate work, crane work, maintenance work, safety monitoring, equipment diagnostics and data operations as the port becomes more intelligent. Workers do not need abstract reassurance that AI will create jobs someday. They need to know where the new work will sit, what training will come before displacement, and whether those roles will be based in Busan rather than purchased from elsewhere.

The third obligation is to treat small manufacturers as operating institutions, not demonstration sites. A factory in Noksan does not become more competitive because a pilot system works during a visit by officials. It becomes more competitive when the system remains in use after the grant period, when workers can maintain it, when managers trust the data, when customers accept the quality records, when downtime falls and when the firm can afford the next upgrade without starting again from zero. If manufacturing AX is to mean more than consulting and installation, Busan should measure the ordinary afterlife of each project: one-year retention, internal staff capacity, recurring maintenance costs, defect rates, delivery reliability and new orders.

The fourth obligation is to turn the Samsung Electro-Mechanics investment into a wider industrial ecosystem. The 15 trillion won plan gives Busan a major AI-infrastructure component anchor, but the city’s broader gain will depend on what grows around it: supplier capability, testing services, process engineering, R&D collaboration, vocational pathways, quality-control jobs and local firms that can participate in the next layer of the value chain. A large investment can lift a city’s profile. Only a connected ecosystem can change its economic structure.

The final obligation is to keep the value chain from slipping away after the announcement. Busan can attract institutions, vendors, platforms, component investment and global attention and still lose the durable value if contracts and operating arrangements are written poorly. Data rights, maintenance authority, technical documentation, local subcontracting, university practice, wage ladders, supplier development and one-year performance audits should not be added later as social benefits. They are the instruments through which a maritime AI strategy becomes a local economy. When they are absent, Busan supplies the site and the story while someone else captures the recurring work.

Busan’s AI bet is therefore not only a technology strategy. It is a test of whether the city can govern the value chain around its own industrial future. The city already has assets many regions try to invent: a major port, a working manufacturing belt, maritime institutions, logistics firms, component production and a real connection to global trade. What it has lacked is control over the higher-value functions attached to those assets. AI will either correct that weakness or deepen it. If Busan uses AI projects and advanced-industry investment to build local maintenance capacity, data governance, technical occupations, supplier ecosystems, maritime services and firms that earn recurring revenue from the systems they operate, the maritime-capital agenda can become more than a political label. It can change the city’s economic position. If it does not, Busan will still get smarter terminals, smarter factories, smarter components and smarter slogans, but the intelligence will live elsewhere.

The work must remain.