LG's AI Graduate School Tests South Korea's New Route to Advanced Tech Talent

Thirty students, one year and problems drawn from the workplace

LG will open South Korea's first company run graduate school devoted to artificial intelligence on September 30, 2025, in Seoul's Magok district. Approved by the Education Ministry in August, the LG AI Graduate School plans to recruit an initial class of 30 master's students in the coming weeks, with teaching scheduled to begin in March 2026. Its degrees will carry the same legal weight as those awarded by conventional universities.

Students will arrive with more than an interest in AI. Each applicant must bring a clearly defined problem from a factory, laboratory or design team, then work toward a solution over three semesters in one year. Graduates are expected to take their results back into daily operations, turning research into improvements that production lines and product teams can measure.

LG's school grew out of an internal program launched by LG AI Research in 2022. Its accreditation comes as major employers compete for AI specialists and try to keep skilled workers from leaving South Korea for technology hubs overseas. The new arrangement offers employees a route to advanced study without leaving their employer.

South Korea's experiment joins two ambitions that do not always move at the same pace: rigorous graduate education and the demand for usable results inside businesses. Companies gain a direct role in developing researchers, while students gain access to industrial data and equipment. Universities and regulators face the task of maintaining academic standards as more training moves inside corporate organizations.

LG will be the first company to test this model under a new legal framework. Other major groups already run internal training or degree programs with universities, making the school part of a wider effort to secure advanced technical skills.

How the law opened the door

A new law that took effect in January 2025 made the graduate school possible. The Special Act on innovation for advanced industry talent allows companies to establish internal graduate schools in fields including semiconductors, artificial intelligence, biotechnology and mobility.

Korean companies could previously operate internal colleges, but those programs were limited to qualifications equivalent to associate or bachelor's degrees. Graduate credentials were awarded by universities or through joint arrangements in which a university conferred the degree.

The new framework allows a company meeting quality standards to operate graduate programs accredited by the state. The Education Ministry describes the system as a way to connect training focused on technology with academic research, while allowing businesses to develop advanced professionals using their own resources.

LG's launch follows several steps, with a further decision needed before it can offer doctoral study:

• 2022: LG AI Research launched the internal program from which the graduate school developed.
• January 2025: The new law took effect, allowing companies to establish graduate schools in advanced industries.
• August 2025: The Education Ministry approved LG's graduate school.
• September 30, 2025: The school is scheduled to open in Seoul's Magok district.
• March 2026: Classes are scheduled to begin for the initial master's cohort.

LG also plans a doctoral program, subject to Education Ministry review. No date has been given for that approval. Employees who enrolled before formal accreditation will still be able to receive recognized credentials, connecting the earlier internal program with the new institution.

Admission starts with a business problem

LG's admissions process will place research proposals at its center, rather than following the usual route of test scores and general statements of interest. Applicants must propose work connected to their current or planned role in the company. Reviewers will assess both ambition and feasibility.

Administrators describe the goal as a practical solution that a business unit can adopt quickly, rather than simply an impressive credential. Students will begin by mapping an existing workflow, defining measures of success, selecting models and running experiments on company systems.

The three semesters are designed to move from theory to experimentation and then deployment, the stage when a model becomes part of a working process. A student might begin with a seminar on the models relevant to a particular problem, build and test prototypes using internal datasets, and finish with a working pipeline ready for operational use.

Results are expected within a year so that the sponsoring unit can adopt and maintain the solution. The schedule follows the pace of product development more closely than the longer timelines common in academic laboratories. That can sharpen focus and speed, but it also requires projects of a manageable size and strong support for preparing and organizing data.

Students working in materials science, manufacturing and product design will have access to datasets and production systems rarely available inside universities. Their work can therefore be tested on actual equipment at scale, with the constraints of a factory, laboratory or product team built into the research from the beginning.

Theory remains part of the promise

Lee Hong lak, co head of LG AI Research, will serve as the school's first dean. The campus will be in Magok, western Seoul, and the curriculum will connect core machine learning and computing concepts with applications across LG's electronics, chemicals, energy and biotechnology businesses.

LG also plans to develop research leaders in generative models, which produce new content; multimodal learning, which works with different forms of information; and causal inference, which examines cause and effect. These subjects sit alongside the practical projects students bring from their workplaces.

Core coursework will include optimization, statistics, model generalization and systems engineering. In accessible terms, those subjects address how models improve, how researchers interpret data, whether a model works beyond its training examples and how the parts of a working system fit together.

Faculty and mentors will include researchers recruited from leading institutions and senior scientists and engineers from LG's AI Research Institute. LG plans a mix of scholars early in their careers and industry veterans, connecting research reviewed by academic peers with the operational limits that matter in factories and data centers.

Students may draw on experimentation that can test many materials quickly, sensor information from smart manufacturing lines and large computing clusters used to train and run models. Access to such resources can be a bottleneck in university research, particularly when the data belong to a company.

Program leaders say theoretical study remains central. The difference is the starting point: students will adapt academic tools to an identified industrial problem, then deliver code and models that a plant, laboratory or product team can take responsibility for maintaining.

A battery researcher describes the change

Shin Dong hwa, a senior researcher at LG Energy Solution, joined the earlier program to move beyond his daily work deploying models. His research challenge was to use AI to accelerate the design of cathode materials at the atomic level.

Shin completed the program earlier in 2025. He said it gave him time to study the foundations of his work and examine modern model architectures in depth:

In the workplace, I rarely had the chance to study theory and mostly focused on applying pre built models. Through the master's program, I was finally able to study AI theory and state of the art model structures in depth, gaining knowledge and ideas I can now use directly in my work.

His experience illustrates the combination LG is seeking: deeper theoretical understanding tied to a problem that matters to one of its businesses. For a researcher already applying AI, the opportunity is to understand and develop the tools more fully, rather than rely mainly on existing models.

Shin also described how access to workplace resources shortened the distance between learning and experimentation:

I can now take ideas from a seminar and test them on real data and hardware within weeks, not months. That accelerates both learning and impact.

For participants such as Shin, advanced study can therefore become part of an existing research career. The company gains work directed at a laboratory's needs, while the employee gains time and structure to investigate methods beyond the immediate demands of daily assignments.

Other employers are building their own pipelines

LG is the first mover under the new law, but South Korea's major technology employers already use several routes to train specialists. SK Hynix, the world's second largest memory maker, has operated SK Hynix University since 2017. It provides short courses and workshops in semiconductor skills rather than awarding degrees.

SK Hynix also sponsors bachelor's programs in semiconductor education at Korea University, Sogang University and Hanyang University. Samsung Electronics operates the Samsung Semiconductor Institute of Technology, an accredited internal university established in 1989. Its master's and doctoral tracks run in partnership with Sungkyunkwan University, combining corporate participation with university involvement in graduate education.

Hyundai Motor has explored an internal graduate school model for mobility and software. Other companies are widening recruitment and strengthening campus partnerships alongside their own training programs.

Kakao recently announced recruitment across its affiliates aimed at professionals described as AI native. Naver Cloud formed a consortium with leading universities to place graduate students in research residencies connected to a national foundation model project. Foundation models provide a base that can be adapted for different AI tasks.

KT has hired experienced AI practitioners while expanding its work with global vendors. LG CNS is increasing internal training and certification, aiming to build a workforce of roughly 1,000 AI experts.

These approaches serve different needs. Short courses update existing skills, university partnerships support degree study, and research residencies connect students with live projects. LG's new school adds another route by placing an accredited graduate institution inside the company itself.

Speed brings a debate about academic independence

The Education Ministry argues that accredited corporate programs can shorten the distance between learning and deployment. Supporters point to fields where models change quickly and where data, computing resources and specialist knowledge often determine whether research becomes a working product.

Supporters also see a way to slow the departure of skilled workers overseas. Employees can pursue advanced training while retaining a role inside their company, giving high performers a clearer path for development at home.

Academic leaders welcome additional investment in training, but warn against organizing graduate education entirely around narrowly defined deliverables. Lee Kyoung jun, a professor of AI and business at Kyung Hee University, contrasted the university mission with the pressure of company priorities:

A corporate graduate school can deliver strong, one time lectures from practitioners. But sustaining a rigorous curriculum is different. Universities exist for independent inquiry. Companies risk narrowing the focus too much to their immediate needs.

Public research budgets are another concern. Universities rely heavily on competitive government grants. Competition from corporate institutes for the same funding pools could put further pressure on already tight budgets.

Many companies already support university research through joint laboratories and degrees supervised together. Supporters of the new model argue that both routes can grow: businesses can finance internal schools while continuing to fund open research on university campuses.

Choi Eun hee, a director general in the Education Ministry, said the schools are intended to complement traditional graduate programs and encourage cooperation:

This system allows companies to cultivate advanced talent directly, using their own resources. The approval of LG's graduate school sets a new paradigm for human resource development in high tech industries and is expected to further stimulate industry academia cooperation.

What changes for universities and student researchers

South Korea's universities remain central to fundamental research. Government funded AI graduate schools at KAIST, Korea University and Sungkyunkwan University launched in 2019, building faculty, laboratories and degree programs that serve the wider economy.

Those institutions train people who can move into startups, public research institutes and different companies without being tied to one employer. Their responsibilities also include open publication and research that may take a long time to produce practical results.

Corporate graduate programs are likely to concentrate more on applied research. That focus can produce rapid returns and valuable intellectual property. It also makes publication rules, data sharing and the ability to carry research experience across employers important parts of the educational arrangement.

Regulators will need to monitor the rigor of coursework and the fair assessment of research contributions. Rules on transferring credits between institutions, protecting student researchers, confidentiality and authorship will help determine the model's quality and credibility.

Students may work with data that a business needs to keep private while also producing research assessed for a recognized degree. Clear arrangements for what can be published, who receives credit and how work is evaluated will shape their experience as both employees and researchers.

AI training extends beyond graduate school

South Korea's education effort reaches further than the new corporate model. The government plans to train all teachers in AI technology and is rolling out AI digital textbooks in schools.

The aim is to build understanding before students enter university. By the time they reach technical majors, students should already have a grounding in how modern AI systems work and how to use them responsibly. Other countries are also bringing computing and data science into general education.

Government agencies fund workforce projects that connect companies with university laboratories. National supercomputing and data initiatives seek to reduce barriers for laboratories without access to proprietary infrastructure.

LG's school fits into this broader chain of education and research. Computing capacity, data and knowledge of a particular field are all needed to turn AI research into products and processes that can operate at scale. Training researchers addresses one part of that task, alongside access to the resources needed to test their ideas.

Global competition adds pressure over scale and control

South Korea's corporate graduate school experiment sits within an international contest over AI talent, data, models and deployment. China's large user base and rapid introduction of digital services give it advantages in scale. Western countries are seeking ways to pool resources and align rules so companies and laboratories can train larger models and deploy them safely.

For an economy of South Korea's size, the strategy often combines national investment, close cooperation between industry and universities, and selected partnerships with global vendors.

Governments and companies also discuss digital sovereignty, meaning control over sensitive data, language models and cloud infrastructure. Vendors offer dedicated cloud regions and specialized hardware to support those goals.

Some analysts argue that sovereignty claims can conceal continued dependence when core technologies remain proprietary. Contracts governing access to computing resources and software, rules on data use and commitments to develop local skills will determine how much national capacity these arrangements build.

LG's first cohorts will provide an early test of the education model. Attention will fall on the balance between theory and applied work, the publication record and the quality of projects put into operation inside LG businesses.

Other companies are studying the model while regulators collect feedback from industry and universities. Expansion could create a small network of accredited corporate graduate schools in priority fields such as semiconductors, robotics and biotechnology. Rigorous curricula, transparent governance and graduates able to contribute both to their employers and the wider scientific community will shape that prospect.

Key Points

• LG's AI graduate school received Education Ministry approval in August 2025 and is scheduled to open on September 30 in Seoul's Magok district.
• The initial class will have 30 master's students, with teaching beginning in March 2026 and degrees carrying the same legal weight as university qualifications.
• Students must propose a workplace research problem and work toward a usable AI solution over three semesters in one year.
• A law effective from January 2025 allows companies meeting quality standards to operate accredited graduate schools in advanced industries.
• LG plans to combine theoretical coursework with industrial data, equipment and computing resources. A doctoral program awaits ministry review.
• Supporters expect faster application of research and stronger employee development. Academic concerns include independent inquiry, curriculum quality and competition for public funding.
• Publication rules, student protections, research assessment and continued cooperation with universities will help determine the model's credibility.