On September 30, 2015, POSCO agreed to pay Nippon Steel & Sumitomo Metal 30 billion yen. The Korean steelmaker stood accused of stealing the recipe for grain-oriented electrical steel, a specialty product used inside power transformers. Nippon Steel claimed POSCO got there by recruiting its former employees. Rather than fight to a verdict, POSCO settled, and the two companies dropped every case on three continents.
There is a detail in that story that almost every account of Korea industrial espionage leaves out. Nippon Steel is the company that taught Korea how to make steel in the first place.
Back in 1970, Japanese engineers helped break ground at Pohang. Japanese money, released by the 1965 normalization treaty, paid for a large share of the furnaces. Forty-two years later, the teacher sued the student. And the student wrote a check.
That reversal is the entire history of Korea industrial espionage compressed into one transaction. For roughly thirty years, Korea acquired other people’s technology by every legal and semi-legal means available. Today Korea sits at the front of the pack, and it is writing some of the harshest technology-theft laws in Asia to stop the next country from doing exactly what it did.
Foreign coverage of Korea industrial espionage usually starts and stops at “China is stealing Korean chip technology.” That version is true, but it is also half a sentence. The other half explains why Korean officials understand the threat so precisely. They are watching a method they invented.
Korea Industrial Espionage By The Numbers
Consider the data first, because the data is why Korea industrial espionage became a political emergency rather than a legal footnote.
In 2025, Korean police investigated 179 industrial technology leak cases and arrested 378 people. The year before, the same agency handled 123 cases and 267 arrests. That works out to a jump of more than 40%, according to the National Police Agency’s National Investigation Headquarters.
Thirty-three of those cases involved technology crossing a border. China was the destination in 18 of them, or 54.5%. Vietnam accounted for 12.1%, while Indonesia and the United States tied at 9.1% each. By sector, semiconductors led at 15.2%, followed by displays at 12.1%, secondary batteries at 9.1%, and shipbuilding at 6%.
Two figures inside that dataset deserve far more attention than they usually get.
First, employees were responsible for 82.7% of cases. Industrial spying in Korea is overwhelmingly an inside job rather than a cyberattack launched from a foreign server. Second, small and medium enterprises absorbed 86.6% of the damage, against 13.4% for large corporations. The public pictures Samsung as the victim. In practice, the typical target is a 40-person supplier in Gyeonggi Province with two patents and no legal department.
Meanwhile, detection of “national core technology” leaks has climbed steeply. There were 9 such cases in 2021. By 2025 there were 33. Government estimates put cumulative economic damage since 2020 at roughly 23 trillion won, or about $17 billion.
One caution belongs with those figures. Rising detection is not the same as rising crime. Korea added investigators, added statutory categories, and added corporate reporting duties over the same period, so some of the increase reflects better vision rather than worse behavior. Even allowing for that, the direction is unambiguous, and the sectors involved sit at the center of national export earnings.
Public opinion tracks those numbers closely. Surveys cited during the 2026 legislative debate found 90.7% of respondents wanting tougher penalties, while 92.5% described technology leaks as a serious threat to the national economy. Consequently, politicians moved.
The Law Korea Rewrote After Seventy-Three Years
On February 26, 2026, the National Assembly did something it had avoided since 1953.
Lawmakers amended Article 98 of the Criminal Act. Until then, espionage in Korea legally meant espionage for an enemy state, which in practice meant North Korea and nothing else. Handing a semiconductor process to a Chinese competitor was therefore a trade-secret matter, not a national-security crime. The revised article replaces “enemy state” with “foreign country,” and offenders now face a minimum of three years in prison. Reinforced technology-leak provisions take effect in September 2026.
Notably, “foreign country” includes allies. An engineer who moves proprietary process data to an American or European employer now falls inside the same statute as one who moves it to Shenzhen. Korean defense exporters have already flagged this as a problem, because their business model depends on transferring sensitive know-how to partner governments.
At the same time, enforcement capacity expanded. On June 29, 2026, the Korean Intellectual Property Office launched a dedicated technology-leak police division, raising its special judicial police force from 27 officers to 61. The new unit staffs 21 investigators drawn from patent examiners, lawyers, and engineering PhDs. Its commissioner has said publicly that he wants at least 100.
So the Korea industrial espionage regime now has three pillars: an espionage statute, a specialist police force, and a fine ceiling of 6.5 billion won. Obviously the country decided this mattered enormously. Less obviously, the technology Korea is defending arrived through channels that look uncomfortably familiar.
How Korea Learned To Build A Chip
Here is the part that rarely survives into English coverage of Korean technology theft.
Samsung’s semiconductor business did not begin with a laboratory breakthrough. Instead, it began in 1976 with a project team, a screwdriver, and a Panasonic microwave oven. The team took the oven apart, studied it, and shipped a Samsung version by 1978. Three years later they repeated the exercise with a VCR. One academic assessment of that period is blunt about the method’s limits: the company “showed no improvement in creative development unless a similar sample or manual was available.”
Memory chips followed the same logic, though with paperwork attached. In June 1983, Samsung licensed a 64K DRAM design from Micron Technology, a mid-sized American producer. Before settling on Micron, Samsung had approached Texas Instruments, Motorola, NEC, and Toshiba. For manufacturing process knowledge, Sharp supplied access to 16K SRAM and 256K ROM production. Samsung engineers subsequently reverse-engineered the Micron design in detail while developing the next generation, as documented in a Berkeley study of the company’s technological capabilities.
People came next. Samsung recruited Korean-American engineers with DRAM experience out of US firms, then built a facility where that experience already lived. Samsung Semiconductor Inc. opened in Silicon Valley and began wafer production in 1985 with 300 engineers. The 256K DRAM was then developed by two teams running in parallel, one in California and one in Korea.
None of this was illegal. That is precisely the point. Licensing, reverse engineering, and hiring experienced staff away from rivals formed the standard toolkit of every late industrializer. Korea simply used the toolkit harder and faster than anyone else.
Yet those same three activities now describe most of the conduct that Korea’s technology leak statutes are designed to punish. Reverse-engineer a competitor’s product today and you may simply be doing product research. Reverse-engineer a designated national core technology, however, and prosecutors have a statute waiting. Almost every element of a modern Korea industrial espionage case has a direct ancestor in the 1976 microwave project or the 1983 Micron contract. What changed was not the behavior. What changed was who owns the thing worth copying.
The Pony Was Assembled From Other Countries
Automobiles tell an even more literal version of the story.
In 1974, Hyundai hired George Turnbull, formerly managing director of Austin and Morris at British Leyland. Turnbull then brought five British engineers with him: a body designer, two engineers, a chassis specialist, and a chief development engineer. Korea did not have those skills yet. So Korea bought the men who did.
The resulting car borrowed nearly everything else. Engines and transmissions came from Mitsubishi. The platform derived from the Mitsubishi Lancer. Certain parts carried over from the Ford Cortina that Hyundai already assembled under license. Italdesign Giugiaro handled the styling.
The Pony launched in December 1975 as Korea’s first mass-produced car. Exports began the following year to Chile, Argentina, Colombia, Ecuador, and Egypt. By 1984, the second generation sold 25,123 units annually in Canada against an initial forecast of 5,000.
British management, Japanese mechanicals, Italian design, Korean assembly. Nine years later it was beating its own projections in North America by a factor of five.
Pohang Was Built With Japan’s Money And Japan’s Engineers
POSCO’s founding follows the same template, except the state did the negotiating.
The government created Pohang Iron and Steel in 1968 and installed Park Tae-joon, a retired general, as president. Financing arrived from Japan under the 1965 normalization settlement: roughly $119 million in Japanese government grants and loans, plus $54 million in credit from the Export-Import Bank of Japan. Technical assistance came from Nippon Steel and other Japanese firms.
Construction began on April 1, 1970. The plant was dedicated on July 3, 1973, with initial capacity of 1.03 million metric tons a year. Within a generation, POSCO ranked among the most efficient steelmakers on earth. Japanese executives, meanwhile, had begun describing the transfer as the worst commercial decision their industry ever made.
That resentment is the missing context for what happened in 2012, and it is why any honest Korea industrial espionage timeline has to run in both directions at once.
The Three Boomerangs Of Korean Technology Theft
Three cases show the reversal cleanly, since in each one the party that supplied the knowledge later went to court over it. Together they form the spine of the Korea industrial espionage story.
Boomerang one: Nippon Steel and POSCO. Taught Korea to make steel from 1968 onward. Sued POSCO in 2012 over grain-oriented electrical steel, alleging POSCO obtained the process by recruiting former Nippon Steel staff. Settled in 2015 for 30 billion yen. The alleged method — hire the other side’s experienced people — is exactly the method Hyundai had used on British Leyland in 1974.
Boomerang two: Micron, Samsung, and China. Licensed Samsung its first DRAM design in 1983. Four decades later, Samsung became the target instead. In 2023, prosecutors charged a 65-year-old former Samsung and SK hynix executive with attempting to replicate an entire Samsung fabrication plant in Chengdu. Investigators described it as an effort to copy a whole chip plant rather than a single process. He recruited roughly 200 engineers from Samsung and SK hynix, then raised over $350 million from Chinese investors after a $6.2 billion approach to a Taiwanese backer failed. Samsung’s estimated loss was $230 million. Separately, another former Samsung executive received a finalized sentence of six years and four months for transferring technology to China.
Boomerang three: the poaching model comes home. Between 2006 and 2009, Kolon Industries conspired with former DuPont employees to obtain the trade secrets behind Kevlar, aiming to improve its own para-aramid fiber. In April 2015, Kolon pleaded guilty to conspiracy to convert trade secrets. The company paid an $85 million criminal fine plus $275 million in restitution. One DuPont engineer who cooperated served 18 months in prison.
Look closely at the mechanism in all three. Nobody hacked a server. In every case, knowledge walked out inside a human being who changed employers. Furthermore, that is still how it works today, since employees account for 82.7% of Korean cases. The tool has not changed in fifty years. Only the direction of travel has.
Korea is far from unique here, and saying so plainly matters. The United States industrialized the same way. Samuel Slater memorized British textile machinery and rebuilt it in Rhode Island in 1790, earning the nickname “Slater the Traitor” back home. Francis Cabot Lowell toured British mills in 1810, committed the power loom to memory, then reconstructed it in Massachusetts — an episode the Smithsonian describes bluntly as industrial espionage. Britain had banned the export of both the machines and the mechanics. It made no difference.
Every follower economy runs this playbook. Korea ran it exceptionally well. The awkward moment arrives when the follower becomes the leader and has to switch sides of the argument mid-career.
Recognizing that pattern does not excuse anything. Instead, it explains why the current Korea industrial espionage debate feels so charged inside the country. Officials are not defending an abstraction. They are defending capabilities their own generation watched being assembled from imported parts, and they know precisely how quickly such an assembly can be repeated.
The Talent War No Korea Industrial Espionage Law Can Stop
Criminal statutes assume a villain. However, the mechanism draining Korean know-how in 2026 is mostly not criminal at all. It is a hiring market, and it sits almost entirely outside the reach of any Korea industrial espionage prosecution.
Foreign firms have priced Korean chip veterans aggressively. Nvidia has offered up to roughly $258,800 a year for HBM engineers with eight or more years of experience. Broadcom has gone to $146,000, Micron to $116,000. China’s CXMT reportedly approaches Korean engineers individually with offers around triple their current pay. Elon Musk personally sought Korean process engineers in February and May 2026. Intel hired a former SK hynix chief executive as a senior vice president. TSMC and Japan’s JASM have posted openings aimed squarely at Korean university graduates.
The supply side makes those offers effective. Korea’s semiconductor industry association projects demand of 304,000 workers by 2031, against roughly 5,000 new entrants each year. A shortfall near 50,000 looks structurally unavoidable.
Domestic competition compounds the pressure. More than 200 Samsung union members moved to SK hynix in a single four-month stretch through April 2026, drawn partly by bonus gaps. SK hynix paid out around $476,000 per employee, while Samsung’s memory division paid roughly $400,000 and its foundry division about $135,000. A June 2026 survey found 81.5% of Samsung foundry staff wanting to leave within two years.
Then came the part that shows how far the legal logic has traveled. In July 2026, Samsung obtained an injunction blocking two former chip workers from joining SK hynix for 18 months, citing national security concerns around core semiconductor technology. Both companies are Korean. National-security language built to stop leaks to Beijing is now being aimed across Seoul.
Recruitment has also become an attack surface in its own right. Korean firms already face fabricated applicants and social-engineering campaigns, a pattern our reporting on North Korea’s crypto theft operation traced in detail: the fastest route into a company is increasingly the hiring pipeline rather than the firewall. Engineers themselves keep moving as bonus structures and HBM hiring diverge between the two Korean giants.
Why Korea Industrial Espionage Law Is Not Deterring Anyone Yet
New statutes make headlines. Sentences make deterrence. On that second measure, Korea industrial espionage enforcement still has a visible gap.
Despite a fine ceiling of 6.5 billion won — already raised tenfold — roughly one in two people convicted of domestic industrial technology offenses receives a suspended sentence. In other words, half of all convictions produce no prison time whatsoever. Against a payoff running into hundreds of millions of dollars, a suspended sentence is not a deterrent. Rather, it is a business expense.
The contrast with Taiwan is instructive. When Taiwanese prosecutors pursued a case involving TSMC’s 2-nanometer process, the defendant received ten years. Taiwan applied the national core technology provision of its National Security Act for the first time to get there. Korea’s most severe finalized semiconductor sentence to date runs six years and four months.
Two further bills are pending in the National Assembly. One would raise both prison terms and fines again. The other would make punitive damages default to five times actual losses instead of capping them at that level.
Nevertheless, sentencing addresses only part of the problem. The structural weakness sits somewhere else entirely.
Recall that 86.6% of victims are small and medium enterprises. These are the firms supplying etching gases, deposition equipment parts, display materials, and battery separators to the giants. They rarely have the security infrastructure, the retention budgets, or the litigation capacity to defend a process that took fifteen years to refine. A Chinese competitor does not need to breach Samsung. It needs to hire four people from a subcontractor in Cheonan.
Criminal law can punish document theft. Even so, it cannot make an engineer forget what he knows, and it cannot stop him from selling that knowledge lawfully by accepting a job. That limitation is not a drafting flaw. It is the same gap Korea itself exploited for three decades.
What Korea Industrial Espionage Rules Mean For Investors
For foreign readers with money or careers exposed to Korea, the new Korea industrial espionage framework carries several practical implications.
Foreign engineers face a wider statute than before. The espionage revision covers foreign countries generally, allies included. Anyone employed by a Korean technology firm should assume that process documentation, yield data, and equipment specifications now carry criminal weight regardless of destination. Practices considered normal elsewhere — keeping a portfolio of past work, carrying reference material to a new employer — are genuinely risky here.
Non-compete enforcement has hardened. Samsung’s 18-month injunction against two engineers moving to a domestic rival signals what courts will accept. Before signing with a Korean chip, display, or battery employer, foreign candidates should read the restrictive covenants carefully and assume they will be enforced.
Cross-border deals will move more slowly. Designation as national core technology triggers government review of foreign acquisitions, joint ventures, and licensing arrangements. As the designated list expands, so does the approval surface. Investors evaluating Korean materials, equipment, or battery suppliers should price regulatory clearance into their timelines rather than treating it as a formality. This adds friction to a market already flush with capital, as our look at the cash pile Korean chipmakers cannot deploy explains.
Defense exporters face a real conflict. Korean arms sales depend on offset agreements that transfer manufacturing know-how to buyer nations. An espionage law treating any foreign transfer as potentially criminal sits awkwardly against a national policy of selling weapons systems with technology packages attached. Expect litigation to clarify where the line eventually falls.
Supplier due diligence matters more than headline risk. Because small firms absorb most of the damage, exposure in a Korean tech portfolio usually sits several layers beneath the listed company. Ask about turnover among senior process engineers. Ask whether critical know-how is documented or simply lives in three people’s heads.
Chinese competition is a schedule problem, not a capability question. Leaked processes compress timelines. They rarely deliver yield, and yield is where Korean manufacturing advantage actually lives. The useful question is therefore not whether Chinese rivals obtain the technology. It is how many years they need to run it profitably, and whether Korean firms can stay ahead of that clock. Our comparison of the four major chip powers covers where each one currently stands.
Graduating Costs Something
Korea’s industrial ascent is genuinely one of the great stories of the twentieth century. A country with no steel, no cars, and no chips in 1960 now builds all three at world scale.
The mechanism behind that ascent, however, is not the one usually celebrated in retrospectives. It was not primarily invention. Instead, it was acquisition, absorption, and relentless improvement of things other people invented first — the same pattern that built the conglomerates now defining Korean business and, eventually, the technologies Korea is known for today.
Followers want open borders for knowledge. Leaders want fences. Korea spent thirty years on one side of that argument and now finds itself, rather suddenly, on the other.
The 2026 laws are the sound of a country changing chairs. They will probably work at least partly, because Korea tends to execute administrative projects well. Even so, the deeper irony refuses to disappear. Every measure Korea now deploys against industrial spying is a measure that, applied fifty years ago, would have made modern Korea considerably harder to build.
Ultimately, that is not hypocrisy. Rather, it is what graduating looks like. The bill for a Korea industrial espionage crackdown simply arrives later than the diploma.
