“In a month, a new artificial intelligence (AI) model will clearly outperform current employees. There are no talents in the field who have received perfect AI education and can solve problems instantly. Universities should focus not on mere technical skills but on developing the ability to identify and define real problems in the workplace.”
This statement was made at the '2026 Industry University Education Innovation Forum' held on September 21 at the Korea Press Center in Jung-gu, Seoul, co-hosted by the Ministry of Education and the Korean Accreditation Board for Engineering Education (ABEEK). This remark from a corporate representative at the forefront of AI industry innovation highlights the serious 'job-education mismatch' between academia and industry in the rapidly changing technological landscape.
The forum brought together government officials, university representatives, and industry experts to seek concrete solutions for improving university engineering education in line with the rapidly evolving AI ecosystem. The focus was on whether universities are cultivating 'practical talents' who can effectively control and validate AI by integrating foundational engineering knowledge with business domains, rather than just superficially handling the latest AI tools.
Kim Tae-sung, who chaired the comprehensive discussion, raised fundamental dilemmas regarding AI education in universities from an educational engineering perspective.
“Deans of engineering colleges across the country are deeply contemplating the direction of AI education. While universities have established some foundational AI education systems, they are still at the stage of figuring out how to organically integrate domain knowledge with AI technology,” he noted.
He added, “From an educational engineering perspective, acquiring knowledge is a process of transferring short-term memory to long-term memory through intense and continuous learning. Only when this transfer occurs can critical thinking and creativity emerge. While encouraging students to use AI tools is beneficial, excessive reliance may hinder the accumulation of complete knowledge. We are at a turning point where academia must decide how much AI usage should be allowed and controlled in an era where AI can make judgments, act, and design on its own.”
Industry experts participating in the discussion candidly shared the paradigm shifts they are experiencing.
Kim Yoo-seok, CEO of Systemran and President of the Korean Society of Information Engineering, stated, “Just a decade ago, chatbot development companies would naturally hire computer science graduates, but now there is a trend to prefer psychology graduates who can use AI tools to solve problems through vibe coding. The speed of technological advancement is outpacing the educational and adaptive speed of universities and companies.”
He explained, “In the past, the constraints of engineering were resources or costs, but now the extreme 'acceleration' itself has become a constraint. In fields that are not traditional majors like computer science, we need to teach the ability to identify problems in existing workflows and overlay AI on top of them. Departments like mathematics and statistics should focus on solid foundational science education that supports AI-based technologies rather than chasing the latest trends.”
Similar assessments were echoed in the finance IT sector, which operates under strict regulatory environments. Lee Woon-kyu, Director of Korean Financial IT, remarked, “The introduction of AI in finance is progressing along three lines: risk management, improving internal productivity, and enhancing customer convenience. However, due to stringent financial regulations and security concerns, many closed networks cannot use external commercial LLMs (like Gemini or ChatGPT) indiscriminately. In such environments, solid AI majors who understand regulatory frameworks and can handle internal data have a distinct advantage over students who merely input AI prompts.”
He continued, “In the past, the focus was solely on data acquisition, but recently the finance sector has shifted to a 'people and problem-centered' approach, asking 'how to use AI' and 'what practical problems to solve.' University education should also encourage students to discover problems independently and acquire knowledge to solve them within real regulatory contexts.”
Faculty members expressed their challenges in applying knowledge to practice while steering education towards foundational skills and project-based learning.
Jung Joon-ho, a professor in the Computer and AI Department at Dongguk University, stated, “As AI-driven coding automation spreads, we realize that humans must ultimately detect coding errors and make detailed corrections. Starting next year, Dongguk University plans to actively promote a foundational skills enhancement program for lower-year students, focusing on live coding and practical exercises in an environment where AI and external research functions are restricted.”
He also shared insights on the structural limitations of industry-academia collaboration capstone design projects. “In the past, we tackled corporate challenges through year-long projects (like ICIP) that led to internships, but due to increased burdens on companies and mismatches with student preferences, sustainability has become a challenge. We are now restructuring projects to design prototypes in the first semester and either link them to actual startups or deepen the technology in collaboration with research labs in the second semester. We hope companies will trust universities as partners to turn students' bright ideas into real value.”
Concerns from vocational colleges were also shared. Woo Ho-jin, a professor in the Computer Software Department at Yeonsung University, noted, “Vocational colleges excel in practical employment capabilities, and we have rapidly introduced AI literacy education for all students through university innovation support projects and the ANCHOR program. However, it is nearly impossible to align the diverse needs of over 80 different departments (including arts, physical education, and aviation) with a single AI education program.”
He added, “Our university is focusing on 'cloud computing' as a specialized field through the New Industry Specialized Leading Vocational College Support Project, quickly curriculumizing practical cutting-edge technologies like big data and AI in conjunction with harness engineering. However, the speed of technological advancement in the industry is far ahead, and the adaptation speed of academic restructuring and faculty is struggling to keep up.”
Before the comprehensive discussion, Kim Beom-soo, a director at PwC Consulting (formerly of KT), clearly outlined the paradigm shift in the AI technology ecosystem and the direction university education should take.
Kim stated, “AI has evolved beyond a mere productivity enhancement tool to become an operating system for businesses and a subject that replaces and executes tasks. We must pay attention to four major technological trends driving corporate execution: 'autonomous AI,' 'physical AI,' 'edge AI,' and 'vibe security.'”
He emphasized, “Autonomous AI focuses on formalizing tacit knowledge in the field and designing knowledge and control structures to optimize massive 'token costs.' Physical AI, which involves deploying robots in manufacturing, also requires validated extendable ready data through digital twins. Therefore, universities must cultivate engineers who can logically instruct AI, control it, and convert field actions into data.”
Meanwhile, specialized sessions exploring the latest trends across various industries received positive feedback. Cha Hwan-joo, Vice President of the Korean Society of Information Engineering (currently at SK), presented an analysis of the current state of university AI curricula from an industry perspective, diagnosing the mismatch between university curricula and industry job flows. He proposed a 'One-Team' system involving schools, professors, companies, and the government, along with three survival strategies for students: strengthening fundamentals, domain integration, and trust and ethics.
Kim Byung-ik, a director at Modu Solution, introduced a case study on AI for automatic generation of elevator design blueprints, stating, “Companies now urgently need fusion talents who can orchestrate RAG (retrieval-augmented generation) and agents based on traditional engineering domain knowledge, rather than just simple coders.”
Jo Sook-hyang, a director at Mirae I&T, emphasized the need for solid foundational knowledge in software, such as data structures and memory management, along with robust industry domain knowledge in finance and logistics to filter out AI hallucinations and optimize complex corporate environments.
In his opening remarks, Kim Woo-seung, President of ABEEK, highlighted the importance of thorough 'quality control' and 'engineering accreditation' in university AI education by citing innovative examples from global prestigious universities.
Kim noted, “Carnegie Mellon University (CMU), which opened the world's first AI undergraduate program, has implemented rigorous 'quality control' measures. With a total of 735 students in the computer science department, 30-35 students are selected for the AI undergraduate program in their sophomore year, resulting in about 100 students participating in the AI degree program in their second, third, and fourth years.”
He continued, “As domestic universities accelerate quantitative expansion with a goal of attracting 300,000 international students, it is essential to ensure the quality of education through an engineering accreditation system to produce verified talents that companies can trust and hire. I intend to actively propose this to the Ministry of Education's higher education policy direction.”
In her congratulatory address, Lee Joo-hee, a university support officer at the Ministry of Education, stated, “In line with the era of superintelligence, the government is implementing regionally tailored policies centered on 'five poles and three special' strategic industries, with the core being a major transformation in university education to timely cultivate the talents needed by the industry. We will spare no effort in providing financial and institutional support to ensure that universities and industries can communicate closely.”
Concluding the forum, Yoo Hee-jin, head of the Ministry of Education's Industry-Academia Cooperation and Startup Support Division, remarked, “AI and advanced technologies are essential foundational infrastructures for higher education and regional innovation ecosystems. We will continue to expand integrated policies that universities and companies can feel through various communication channels, such as the upcoming Industry-Academia Cooperation Expo at KINTEX starting September 30.”
* This article has been translated by AI.
Copyright ⓒ Aju Press All rights reserved.