On a February morning in 2026, the chairman of one of Korea’s largest construction materials groups walked out of a district court in Uijeongbu, cleared of all charges. Three of his workers had been buried alive in a collapsing quarry four years earlier. His case was literally Case No. 1 under the Korea serious accidents act — the very first prosecution brought under a law written to make men like him personally, criminally responsible when workers die. He was acquitted.
Meanwhile, roughly 1.7 Korean workers were dying on the job that same day. They had done so every day since the law took effect. That tension sits at the heart of one of the most watched legal experiments in the industrialized world. Korea did something almost no other country has done. It wrote a statute that sends the person at the very top of a company to prison when a worker dies. Four years later, however, the results look nothing like what either side expected.
Here is what the numbers actually show, and why every foreign executive running a Korean subsidiary should understand them.
The Serious Accidents Punishment Act — 중대재해처벌법, or jungdae-jaehae-cheobeolbeop — passed the National Assembly in January 2021 and took effect on January 27, 2022. Initially, it applied only to workplaces with 50 or more employees. Then, on January 27, 2024, coverage expanded to businesses with as few as five workers. As a result, roughly 830,000 additional small firms came under the law overnight.
The mechanism is deceptively simple. Most industrial safety regimes punish the company, the site manager, or the safety officer. By contrast, this Korean workplace safety law reaches past all of them to a category it calls the “management-responsible person” — in practice, the CEO or, in some readings, the group chairman.
If a worker dies, that person faces at least one year in prison, or a fine of up to ₩1 billion (about US$720,000). The corporation itself can be fined up to ₩5 billion. Furthermore, victims can pursue punitive damages of up to five times actual losses. For serious injuries or occupational illness clusters, the ceiling drops to seven years’ imprisonment.
Crucially, the duty is not “don’t cause an accident.” Instead, it is a duty to build and maintain a safety management system. That means risk assessments, budget allocation, a designated safety organization, subcontractor oversight, and documented follow-up. You can read the official English translation of the Act through the Korea Legislation Research Institute.
Legally speaking, this is closer to the UK’s Corporate Manslaughter and Corporate Homicide Act than to ordinary workplace regulation. Yet Britain prosecutes the organization. Korea, by contrast, prosecutes a named human being.
Foreign readers usually ask the same question first: why would any legislature go this far?
The short answer is Kim Yong-kyun. In December 2018, the 24-year-old contract worker was crushed to death by a coal conveyor belt at a thermal power plant in Taean. He was working alone, at night, on equipment that should have required two people. His death became a national story partly because of one detail. His mother campaigned for years afterward. Through her, the public learned how routinely the most dangerous work in Korea had been pushed onto subcontractors.
That outsourcing pattern even earned a name — 위험의 외주화, “the outsourcing of danger.” Consequently, the reform debate shifted. Reformers argued that punishing a subcontractor’s foreman changed nothing, because the economics that created the risk were decided several corporate layers above him.
Korea also carried an uncomfortable statistic into the debate. For years, its fatal occupational injury rate ranked among the worst in the OECD, despite an economy that had long since joined the rich world. Although workplace deaths had been falling slowly, they had not fallen nearly as fast as Korea’s GDP had risen.
Business groups fought the bill hard, and they still do. Nevertheless, in the political climate that followed Kim’s death, the law passed with unusual speed.
A complete audit of the law’s first four years, covering 2022 through 2025, produced a figure that has since defined the debate.
2,436 workers died in accidents subject to investigation over those four years. That works out to roughly 1.7 deaths every single day, holidays included. The deaths occurred across 1,990 separate workplaces. Notably, 156 of those workplaces recorded a fatal accident more than once.
The industry split is stark:
| Sector | Deaths (2022–2025) | Share |
|---|---|---|
| Construction | 1,205 | 49.5% |
| Manufacturing | 679 | 27.9% |
| All other sectors | 552 | 22.7% |
In other words, half of Korea’s workplace deaths happen on building sites. Anyone who has watched Seoul’s skyline churn will not find that surprising, but the scale still is.
The trend line is arguably the worst part. In 2025, according to Ministry of Employment and Labor figures, 605 workers died in investigated accidents — up from 589 the previous year. Despite four years of criminal exposure at the top, in other words, the number went up. The government’s own fatal accident rate target for 2026 sits at 0.37 per 10,000 workers, barely below the 0.39 recorded in 2025.
The public sector is not exempt either. Korea Electric Power recorded 12 deaths over the four years. Korea Railroad recorded nine. The Korea Forest Service recorded five.
Now for the part that makes the Korea serious accidents act genuinely strange.
By March 2026, Korean courts had handed down 101 first-instance verdicts under the Act. Of those:
Put differently, roughly one defendant in seventeen actually went to prison under a law whose entire premise is that executives should go to prison. Moreover, the average custodial term among those convicted came to about one year and one month — the statutory minimum, essentially. The average corporate fine landed near ₩72.8 million (roughly US$52,000), once one outlier case was excluded.
The acquittal rate deserves attention too. At 10.7% among individual defendants, it runs more than three times Korea’s general criminal acquittal rate of about 3.1%. For a country where prosecutors almost never lose, that gap is remarkable. It tells you something specific: judges are repeatedly unconvinced that the person in the dock was really the person who controlled safety.
The first sentence ever upheld at the Supreme Court involved a mid-sized steel company, Hanguk Steel. Its CEO received one year. Law firms tracking the Korea serious accidents act, including Kim & Chang, treat that ruling as the closest thing to settled doctrine so far.
Here is the finding that reframes everything above: about 83% of indictments under the Act have targeted small and mid-sized enterprises.
That outcome is close to the inverse of the law’s stated purpose. The Act was written because reformers believed conglomerate chairmen were insulated from consequences. Instead, the people most often sitting in the defendant’s chair run companies with a few dozen employees.
The reason is structural rather than conspiratorial. In a small firm, the owner-CEO signs the safety budget, hires the safety officer, and walks the site personally. Therefore, proving that he was the “management-responsible person” is straightforward. In a chaebol subsidiary, by contrast, those same decisions pass through a safety division, a compliance committee, a subsidiary CEO, and a holding company. Each layer creates reasonable doubt.
There is a second effect worth naming. Small Korean firms were already under severe margin pressure, as anyone following Korea’s small business crisis will recognize. For them, the Korean workplace safety law arrived as one more compliance cost. Adding criminal exposure to an owner who cannot afford a dedicated safety team does not automatically produce safety. Sometimes it simply produces paperwork.
Labor advocates counter that this is an enforcement failure, not a design flaw. Business groups counter that it proves the law is unworkable. Remarkably, both sides cite the same statistics.
Five Korean companies recorded a fatal accident in every single one of the law’s first four years:
Daewoo E&C stands out even in that group. Across four years, it recorded 14 separate serious accidents that killed 15 workers — three in 2022, two in 2023, seven in 2024, and three in 2025.
Meanwhile, among Korea’s ten largest construction firms, the number of Supreme Court convictions finalized under the Korea serious accidents act stands at zero. Of the 17 construction-sector rulings issued so far, 15 produced convictions. Those convictions, however, clustered among smaller contractors.
Why the gap? Partly, it reflects the layering problem described above. Partly, it reflects resources: a major contractor can field a defense team that a 40-person subcontractor cannot. And partly, it reflects Korea’s subcontracting chain itself, where a single high-rise project may involve dozens of firms and a correspondingly diffuse chain of responsibility.
For foreign investors reading Korean construction and industrial names, this matters in a practical way. The headline legal risk sits with the top-tier contractor. The realized legal risk, so far, sits further down the chain. That asymmetry also colors Korea’s broader labor law reform debate, where subcontracted workers and their bargaining rights have become the central fight.
Two cases explain the current mood better than any statistic.
On June 24, 2024, a lithium battery plant in Hwaseong, south of Seoul, caught fire. Within roughly fifteen seconds of exposure, toxic smoke from thousands of burning cells rendered the factory floor unsurvivable. Twenty-three workers died. Seventeen of them were Chinese nationals; one was Laotian; seventeen were women.
The Hwaseong battery factory fire was the deadliest industrial disaster in Korea in decades. Moreover, it landed squarely on migrant workers who had received safety training in a language many could not follow. For context on how central foreign labor has become to Korean industry, see our reporting on Korea’s foreign caregiver program and work visa routes into Korea.
In September 2025, Aricell’s CEO received 15 years — by far the heaviest sentence ever imposed under the Act. Then, on April 22, 2026, an appeals court cut it to four years. His son, the operations chief, went from 15 years to seven. The court cited settlements reached with every bereaved family. Furthermore, it concluded that the company had not completely abandoned safety — even though a smaller explosion two days earlier had gone essentially ignored. No ruling has done more to shape public feeling about the industrial accident law.
Bereaved families reacted in the courtroom. “Twenty-three people died,” one relative said. “Is this the law?”
The second case ran the opposite direction. On January 29, 2022 — two days after the Act took effect — a quarry face collapsed in Yangju, burying three workers. It became prosecution Case No. 1.
Four years later, on February 10, 2026, the Uijeongbu District Court acquitted Sampyo Group chairman Chung Do-won and the group’s former CEO. The company itself was fined ₩100 million under separate occupational safety legislation. The court’s reasoning was narrow but consequential: receiving safety reports, it held, does not by itself make someone the management-responsible person. What matters is whether that individual could concretely direct and supervise safety measures.
Labor groups called it a gutting of the statute’s purpose. Corporate counsel called it long-overdue clarity. Either way, the ruling reshaped how every Korean group now structures its safety governance.
If courts will acquit a chairman who merely “received reports,” then the obvious corporate strategy writes itself. Appoint someone else.
Enter the CSO — Chief Safety Officer — a role that barely existed in Korea before 2022 and is now standard at large firms. The theory runs like this. A CSO with genuine budget authority, independent decision-making power, and a direct reporting line becomes the management-responsible person. Consequently, the CEO is shielded.
In practice, courts have treated CSO appointments with visible skepticism. Early rulings held CEOs liable anyway, reasoning that the CSO existed on paper without real authority. More recently, however, at least one court accepted the structure and acquitted a CEO on that basis. Chambers-tracked Korean practices now advise clients that a CSO defense survives only under three conditions. The CSO must control an independent safety budget. In addition, that person must make final calls without CEO sign-off. Finally, documented evidence of both must exist.
Otherwise, the appointment is simply a name on an org chart — and Korean judges have proven quite good at spotting those.
Foreign executives heading Korean subsidiaries tend to underestimate this law, largely because their home jurisdictions have nothing comparable. Below is what compliance actually requires in practice.
1. Document the system, not just the outcome. Prosecutors examine whether a safety management system existed and functioned. Risk assessments, safety budgets, training records, and — critically — evidence that identified hazards were actually fixed all matter. Verbal instructions count for nothing in court.
2. Take subcontractor oversight seriously. Roughly half of fatal accidents involve subcontracted workers. Your duty extends to firms you do not employ, so contract language alone will not protect you. Site-level verification will.
3. Decide the CSO question deliberately. If you appoint one, give that person a real budget and real authority, then document both. A hollow appointment is worse than none, because it looks like evasion.
4. Address language barriers head-on. Aricell made this unavoidable. If your workforce includes migrant workers, safety training in Korean only is now a demonstrable failure of duty. Translated materials and verified comprehension are the standard.
5. Know your first 72 hours. After a serious accident, the Ministry of Employment and Labor and police arrive quickly, and document preservation orders follow. Retain Korean counsel immediately, because early statements shape everything afterward. The Korea Occupational Safety and Health Agency publishes guidance, though enforcement practice moves faster than the manuals.
6. Watch the calendar for adjacent reforms. The Korea serious accidents act does not operate alone. Korea’s rising statutory retirement age is keeping older workers on sites for longer. Meanwhile, heat exposure, cardiac events, and emergency medical response gaps increasingly show up in serious-accident casework.
The law is now in a peculiar political position. Business federations want it softened, arguing that vague duties and criminal exposure have produced compliance theater rather than safer sites. Labor federations want it strengthened, pointing to a 5.94% imprisonment rate as proof that the deterrent never materialized. Both groups, again, argue from identical data.
Three developments will decide the next phase.
First, appellate courts are still converging on what “management-responsible person” means. The Sampyo acquittal and the CSO rulings point toward a narrower definition. Should that hold, the law’s reach contracts sharply.
Second, enforcement capacity remains thin. Korea’s labor inspectorate is small relative to the number of covered workplaces. That shortfall is exactly why indictments concentrate where causation is easiest to prove.
Third, disasters keep resetting the politics. On March 20, 2026, a fire at an auto parts plant in Daejeon — a supplier of engine valves to Hyundai and Kia — killed 14 workers and injured about 60. Each such event pushes the debate back toward enforcement, regardless of where the legal doctrine had been drifting.
For now, the summary is this. Korea built the most aggressive executive liability regime in the developed world. Yet workplace deaths have not fallen. Whether that proves the Korea serious accidents act failed, or merely proves it was never enforced as written, is the question Korean courts will answer next. Foreign companies operating here should assume the answer lands closer to “enforced more consistently” than “quietly repealed.” After all, 2,436 deaths is a number no legislature walks away from.
Every Chuseok, the country empties out in one direction at once. This year the government…
On September 1, 2026, a cracker at Daesan went quiet. For years it had made…
The Sport That Quietly Took Over Korea's Riverbanks Walk along the Han River at six…
The Seoul Apartment That Triggers a Korea Inheritance Tax Bill Picture a two-bedroom apartment…
The story starts with an envelope. In 2014, a court ordered striking autoworkers to pay…
The Building That Is 99.7% Full Somewhere in Gangnam, there is an office tower with…