Quinn Emanuel founder John Quinn sees AI intensifying patent, antitrust and geopolitical risks for Samsung, SK hynix
The artificial intelligence chip boom is opening a new legal front for South Korean chipmakers, with Samsung Electronics and SK hynix facing a US class action alleging they coordinated cuts in conventional DRAM supply while shifting capacity toward more profitable high-bandwidth memory.
The lawsuit, which also names US memory chipmaker Micron Technology, claims the companies’ production decisions constrained conventional DRAM supply and drove up prices.
John Quinn, founder and chairman of US litigation powerhouse Quinn Emanuel Urquhart & Sullivan, said the case would hinge on evidence that the companies had agreed to act together.
Chipmakers responding independently to the same AI-driven market conditions can make similar production decisions without violating antitrust law, he said.
“An antitrust violation requires an agreement,” Quinn told The Korea Herald in a Zoom interview Thursday.
The case highlights how the AI boom is reshaping legal risks for Korean technology companies. As chipmakers pour resources into HBM and other advanced products, they face growing exposure to antitrust claims, patent disputes and geopolitical tensions.
Patent disputes target AI chips
Patent litigation is another growing concern as increasingly valuable technologies such as HBM become targets of intellectual property disputes in the US.
Much of that litigation comes from nonpracticing entities, or NPEs — companies that own patents without necessarily producing the underlying technology and seek licensing fees or damages from alleged infringers.
Quinn said some NPEs, often called "patent trolls," use the high cost of US litigation itself as leverage. They buy relatively weak patents cheaply and pursue companies that may find it more economical to settle than fight.
“They know that it’s expensive to litigate and that just by bringing a claim they can create some value,” Quinn said. “It’ll be worth it to the defendant to pay something to make them go away.”
For Korean companies, the exposure is substantial. Of 610 patent infringement suits filed against them between 2020 and 2025, 553, or 90.6 percent, were brought in the US, according to Korean intellectual property authorities.
NPEs filed 400 of the 528 cases involving the 10 most frequently sued Korean companies. Samsung alone faced 350 suits, including 259 brought by NPEs.
Patent battles are increasingly reaching chips central to the AI boom. The US International Trade Commission recently opened an investigation into claims by Monolithic 3D that SK hynix and Japan’s Kioxia infringed patents covering HBM and NAND flash technologies.
Samsung, meanwhile, has been locked in a yearslong dispute with Netlist over memory technologies including DRAM and HBM.
Quinn cautioned against viewing Korean companies as uniquely targeted, saying NPE litigation is widespread across the US technology industry.
Caught between Washington and Beijing
For Korean companies, legal risk is also becoming harder to separate from geopolitics, as Washington and Beijing tighten controls over strategic technologies. Korean conglomerates operate major businesses and manufacturing facilities in China while remaining heavily dependent on the US market, leaving them caught between the two powers.
“They have to maintain good terms with China,” Quinn said. “They want to be on good terms with the US. It makes it very challenging.”
Despite those pressures, he said major Korean groups are as sophisticated as their US counterparts in handling global litigation.
But a dispute that threatens a company’s core business cannot be left solely to its legal department, he said.
“If it’s really enterprise-threatening, that’s something that is going to require the attention of the CEO, the C-suite and maybe even the board of directors.”
AI opens new legal fronts
The AI boom is also generating disputes beyond conventional patents and antitrust law.
Quinn pointed to unresolved questions over copyrighted material used to train AI models, patent protection for AI-generated inventions and liability for harmful AI outputs.
“There are a whole host of novel legal issues that AI is introducing,” he said, adding that courts would likely adapt existing legal principles to the emerging technology.
AI is also increasing the value of trade secrets, particularly algorithms and other critical know-how that may not qualify for patent protection.
As engineers move more frequently among semiconductor, battery and AI companies, businesses must be able to show that they actively protected commercially sensitive information, Quinn said.
herim@heraldcorp.com