[Edaily Reporter Kim Hyun-ah ] The AI talent war has spilled over into the courts after LG Corp. AI Research filed separate preliminary injunctions against individuals—one regarding a researcher’s move to Upstage from the Ministry of Science and ICT’s “Independent AI Foundation Model (Dokpamo)” project, and another regarding a former executive’s move to KTCorporation.

As the legal battle continues—with the court dismissing the Upstage case in the first instance and LG Corp.(003550)filing an appeal—AI industry insiders who spoke to reporters generally agreed that LG Corp.’s response was excessive for a large corporation.

So why did LG Corp. go so far as to file a lawsuit over the departure of a junior researcher? LG Corp. maintains that it has never prevented the departure of ordinary researchers and that this measure is an exceptional case intended to protect core R&D assets and trade secrets.

On the other hand, Upstage and KTCorporation(030200)emphasized that they followed standard hiring procedures and that the former employee’s current duties differ from their previous ones. They argue that the information held by the former employee does not constitute trade secrets worthy of legal protection.

Do Large Corporations Prohibit Even Junior Employees from Changing Jobs?…LG Corp.: “Access to Information Is More Important Than Job Title”

LG Corp. AI Research stated that Mr. A, who moved to Upstage, should not be viewed merely as a “junior researcher in his second year of a master’s program.” The company argues that the experiences of success and failure, as well as the trial and error Mr. A accumulated during his on-the-job training, are also important R&D assets.

LG Corp. stated, “While employed, Mr. A received a retention bonus—offered to the top 5% of core research personnel—as well as internal awards and top performance evaluations. Upon his resignation, we offered him a 15% increase in his base salary and an additional retention bonus of 20 million won per year for three years.”

The company added, “We understand that Upstage persuaded him to join the company in time for the start of its post-employment training program, claiming that he would only be eligible for a 4+4 stock option grant (totaling 800 million won) if he joined before April 30, the date of the extraordinary general meeting.”

Upstage refuted LG Corp.’s claims. An Upstage official stated, “The employee applied directly through an open recruitment process, and their compensation package is no different from that of second-year employees at our company.”

The spokesperson continued, “This employee is a junior developer responsible for the practical implementation of AI model post-training,” and emphasized, “Since they are handling a different area of post-training technology than what they worked on at their previous employer, this has absolutely no connection to infringing on another company’s technology or leaking know-how.”

Upstage also clarified that Mr. A was not recruited to work on any specific project or technology. Regarding the fact that Mr. A is listed as a contributor in both the LG Corp. AI Research (K-EXAONE 2.0) and Upstage (Solar Open2) Dokpamo second-round evaluation model technical reports, the company maintains that the technical reports are fully公開 and available for verification.

AI Industry: “Non-Compete Clauses Should Be Handled with Caution”

The AI industry is raising concerns that

non-compete clauses

, which are widely used as a means of securing talent, should be handled with caution.

The CEO of an AI technology company said, “To retain key talent, it is more important to offer compensation commensurate with market value rather than restricting job changes.”

He added, “Given that AI companies funded by the National Growth Fund—such as Rebellion, FuriosaAI, and Upstage—are competing for talent within the same ecosystem by offering high salaries, personnel mobility is a natural occurrence.” This suggests that widespread use of non-compete clauses could stifle talent mobility across the entire domestic AI industry.

An AI expert remarked, “Even if Mr. A participated in the second phase of development for LG Corp. AI Research’s ‘Dokpamo,’ it makes no sense to claim that the project was affected by the departure of a single employee. Excessively restricting the career mobility of AI talent could stifle the domestic AI ecosystem as a whole.”

The point is that since the movement of top talent, entrepreneurship, and technology diffusion are key drivers of growth in the AI industry, non-compete clauses should be applied only in extremely limited circumstances.

Lee Seong-yeop, director of the Center for Technology Law and Policy at Korea University, also said, “Non-compete clauses should be applied sparingly—not only for the sake of individual researchers’ compensation but also to secure national AI talent through the dissemination of knowledge.”

AI Talent Lawsuits in the U.S. as Well… Confidential Information, Not “Job Changes,”

Is

the Key Issue In the United States, too

, the recruitment of talent among AI companies is leading to trade secret disputes. Elon Musk’s xAI has taken legal action, alleging that confidential information was leaked during a former employee’s move to OpenAI and that the competitor attempted to exploit it. In this case as well, the key issues were not the job change itself but whether the former employee actually obtained and used confidential information, and whether the competitor utilized it.

In the dispute involving Apple and OpenAI, the issue centered on whether former employees had accessed internal systems and obtained or utilized confidential materials, such as supplier information. In the U.S., too, disputes are proceeding by specifically examining what information is subject to protection and whether that information was actually transferred or used, rather than focusing solely on the act of moving to a competitor.

In South Korea as well, there was a standoff in 2023 when Naver sent a certified letter to an AI expert who had moved to SKTelecom. At the time, the situation did not escalate into a full-scale lawsuit.

The competition for talent in the AI industry is expected to intensify further. This is because, as telecommunications companies, platforms, and manufacturers alike adopt AI as a key driver of corporate growth, the race to secure talent is spreading beyond industry boundaries.

It is difficult to classify general AI technologies acquired through academic papers or open-source resources—along with a researcher’s experience—as the assets of a specific company. On the other hand, confidential data, experimental results, internal evaluation materials, and specific training methods that companies have accumulated at great expense need to be protected as trade secrets.

In particular, since the inauguration of the Lee Jae-myung administration, competition for talent among companies has intensified as budgets for national AI projects—such as the “Dokpamo” initiative—have expanded. With the government investing massive amounts of national budget and computing resources to foster the AI ecosystem, observers point out that standards are needed to protect corporate R&D investments and trade secrets without excessively restricting the movement of talent—which is central to the nation’s AI competitiveness. This lawsuit is expected to serve as a test case for defining the boundaries of legal protection between the protection of corporate trade secrets and researchers’ freedom to change jobs in the context of national AI projects.