Samsung union lost before the strike even began?

```

Today’s court ruling stripped the upcoming action—called “the biggest strike in semiconductor history”—of its core weapon before it began, at least on the legal front. The short-term risk to the supply chain has dropped significantly, but this labor dispute is far from over, and there’s still one card left to play.

On May 18, just three days before Samsung workers were set to leave the factories, a ruling by the Suwon District Court in South Korea fundamentally changed the nature of the incident—or, at least, made judging the strike much more complicated for the public.

The court partially approved Samsung’s request for an injunction. The core language is: During the strike, the union must not stop, abolish, or obstruct the operation and production activities that should be kept running with “the same level of personnel, working hours, and care obligations as usual.” At first glance, this means workers can announce a strike, but staffing on production lines cannot be reduced, and production must continue as usual. But “the court said so” and “what will actually happen” are not necessarily the same thing.

Korean media reacted in unison. The Korea Economic Daily described the ruling as a “complete victory” for Samsung. Maeil Business News headlined, “Court presses brakes on Samsung’s all-out union strike—Operations to maintain normal manpower.” The Munhwa Ilbo was even more direct: “Essentially, it’s now very hard for the strike to go ahead.” These views have legal grounds, but all are based on a premise yet to be tested: whether the injunction can be effectively enforced.

Where is this card being played?

To understand the ruling, you first need to understand the framework of Korean labor law.

Article 42, Paragraph 2 of Korea’s Trade Union and Labor Relations Adjustment Act stipulates that unions may not stop the operation of “safety protection facilities.” This clause was originally meant for genuine safety equipment—devices and procedures that prevent toxic gas leaks, fires, and protect worker safety. There is universal agreement and clarity about this.

The controversy lies in this: Samsung included the entire semiconductor wafer production process—from chemical management to clean room maintenance—as “safety protection facilities,” asking the court to bar the union from reducing any production-related personnel. The court largely accepted this argument.

Labor law expert and attorney Kim Nam-seok analyzed before the ruling that “If a facility is deemed one where interruption would cause major loss or danger, the court may decide that key personnel must be assigned even during a strike, setting an appropriate scale.” The court’s final move went further than the “appropriate line”—almost the entire production line was included in the protected scope.

Another labor law expert, Um Tae-seop, managing attorney at OHJIMS, had earlier pointed out that since semiconductor processes use many toxic and flammable gases and strong acids/alkalis, classifying them as “safety protection facilities” under Article 42 has some legal rationale. But critics argue: Interpreting “safety” this way is essentially using judicial means to restrict the constitutionally protected right to strike. If every step in semiconductor production can be claimed to result in “major loss if interrupted,” what can’t be included?

What has the union lost?

The reason strikes are bargaining chips is simple: Shutting down hurts management enough to make them give in. Samsung estimates daily shutdown losses at $670 million; the union claims a total of about $20 billion for 18 days. These numbers are the union’s hard leverage at the negotiation table.

The court ruling fundamentally broke this chain of logic.

In numbers, Samsung’s semiconductor unit has about 77,000 employees; the court mandates about 4,000 to 8,000 (5%–10%) maintain normal work during the strike—these are essential positions. On paper, around 40,000 employees could still strike, and Korea Financial News cited analysis that “even if 8,000 can’t join, about 40,000 can walk out.”

But this number hides a structural fact: the 5%–10% required to stay are concentrated at critical nodes in the production line. Wafers don’t stop just because most workers are gone—so long as core positions are filled, the line runs. As early as May 14, Samsung started its “strike contingency plan,” prioritizing personnel reallocation to HBM lines, aiming to keep minimal but effective output with court support.

In other words, even with 40,000 out, lines may still run. The cost to management shifts from “shutdown loss” to “paying zero wages to striking workers”—two issues of vastly different scale.

The union faces another dilemma: violating the injunction means about $72,500 in fines per union per day, and about $7,250 per union leader per day. The sum itself isn’t huge, but violating a court order means sharply rising legal risks and a shift in public opinion. Public support for this strike was already low—several surveys show many believe “having a job at Samsung is already very high-paying, and this strike’s demands are excessive.”

This is an unprecedented ruling

In prior Korean judicial practice, it was almost unheard of to designate the semiconductor production process this widely as “safety protection facilities” and thus restrict strikes.

The ruling’s significance extends beyond Samsung. It has drawn a line at the judicial level for Korea’s semiconductor industry: for sectors recognized as national strategic assets, to what degree can the right to strike be overridden by the need for production continuity? If this ruling stands, similar injunctions will become easier to obtain, and the bargaining power of semiconductor workers could be systematically diminished.

The union’s legal team said after the ruling that they “respect the court’s decision, and will proceed with the planned action on the 21st.” This statement packs two messages: We will not defy the law, but we are not backing down, either. Now, the union faces a choice: keep pressing with a strike that’s much less damaging, or return to the table and try for the next best outcome with their main chip lost.

Supply chain risk repriced

For the global AI hardware supply chain, today’s decision hugely reduced the scenario for short-term disaster.

Previously, the main market worry was: Samsung stops for 18 days, a gap arises in global HBM (High Bandwidth Memory) supply, SK Hynix is at full capacity, and Micron also has no unused lines—no one to fill the gap. At a key moment for Nvidia’s Blackwell chip shipments, HBM supply constraints would directly affect AI datacenter delivery. Analysts were especially worried because late May to early June is when Samsung’s HBM4 yield and output ramp up. Problems now could wipe out the market share Samsung spent three years reclaiming.

That logic needs to be discounted now. If core production staff remain, Samsung’s HBM4 shipments will mostly stay on track; the risk of disruption during this critical window is much lowered.

But two variables remain. First, the real effect of enforcing the injunction. If the union chooses "work-to-rule" rather than a full stoppage, how much will output drop, and how will the court define “violations”? Second, whether the government will invoke its “emergency adjustment power”—if exercised, it can suspend the strike for up to 30 days, but this carries high political cost, and the current administration is in no hurry.

Suppressed tensions are unresolved tensions

The court may have helped Samsung win this round at the table, but the underlying conflict remains unchanged.

The union’s main demand is rooted in a real-world comparison: Rival SK Hynix promised last September to abolish bonus caps, allocate 10% of operating profit to employee bonuses for ten years. Samsung semiconductor workers, with surging AI chip demand, see their peers get that deal, while their own employer offers just one-time subsidies and refuses to write anything into contracts. This psychological account won’t be wiped clean by a court order.

Deeper down, the key question is: If Samsung maintains the current pay structure, it will face ongoing structural pressure from SK Hynix in attracting and retaining top engineers. This isn’t a stock price question today, but it is a variable worth attention in Samsung’s long-term competitiveness narrative.

How do investors see it?

The direct impact of the court ruling is a lower risk premium for Samsung in the short term. The market had already absorbed a round of share price impact—after the strike was announced, Samsung lost about $66 billion in market value in a day, then partially recovered. Today’s judgment adds further support to that rebound.

The old “Samsung stops → Orders shift to SK Hynix and Micron” logic is weakened. There may be fewer emergency orders, but the longer-term structural opportunity of “diversifying suppliers” hasn’t gone away—customers will reassess single-supplier risk after this, and that concern will outlast any single strike.

Nvidia’s earnings call tomorrow is worth particular attention. If Jensen Huang mentions HBM supply stability or risk diversification, that will be a signal of this event’s impact reaching top industry decision-makers—even if the court cooled short-term risk, strategic reassessment on the purchasing side may already be quietly underway.

The day after tomorrow, 47,000 Samsung workers will leave the factories. Production lines will almost certainly keep running, and the global AI memory supply chain probably won’t be interrupted.

But the real story today isn’t the court pulling the brake on a strike. It’s that—for the first time—Korean justice has clearly chosen a side between “production continuity” and “the right to strike.” Who ultimately pays the price for that choice remains to be seen.

Risk warning and disclaimerMarkets have risks, invest cautiously. This article does not constitute personal investment advice and does not consider the particular investment objectives, financial situations, or needs of any individual user. Users should consider whether any opinions, views, or conclusions herein suit their own circumstances. Investment is at your own risk. ```