Universities must fundamentally redesign their approach to education as artificial intelligence reshapes the global workforce, according to education experts and industry leaders speaking at the 2026 Korea Times Global Conference in Seoul. The panelists argued that higher education's survival depends on shifting from knowledge transmission to cultivating human-centered skills AI cannot replicate - critical thinking, ethical judgment, and the ability to define problems rather than just solve them.
Lee Ji-hyun, director of the AI Integrated Talent Development Division at the Ministry of Education, said she previously faced resistance when advocating for this shift. "In the past, when I suggested to a high-ranking official that university education should ultimately return to its essence and focus on human capabilities, I was dismissed," Lee said. "Back then, I was told that in the AI era, universities should focus only on how many graduates they produce, not on building human skills."
Lee emphasized that OECD research identifies social-emotional skills - curiosity, creativity, self-control, empathy, and openness - as critical traits for the future workforce. "These social-emotional skills, like curiosity and creativity, help students ask the right questions, build on what AI produces and avoid becoming overly dependent on technology," she said. "Universities must become frontier places where students can truly build these human capabilities."
Rethinking assessment and the role of failure
Gregory C. Hill, chief administrative officer at the University of Utah Asia Campus, challenged universities to abandon traditional testing methods that measure knowledge recall. "We have to redesign our assessments to make students better at reasoning and thinking through problems rather than relying on historical assessment tools like the SAT or LSAT that merely test knowledge regurgitation," Hill said. "You simply cannot outsource character as a function of AI training."
Lee Jae-wook, director of the AI Institute at Seoul National University, pointed to the widening productivity gap between AI users based on their ability to interact with the technology. "The performance of AI is developing so much faster than human capabilities, creating a massive productivity gap between users," he said. "Interaction with AI depends heavily on a person's experience and level of abstraction. Universities play a vital role in building these foundational blocks of knowledge and developing soft skills, leadership and human interaction through community."
Lee added that universities must become spaces where students can safely make mistakes. "Because society is changing so fast and there is no consensus yet on the right path, we must go through continuous loops of trial and error," he said. "Universities must serve as fertile soil for these mistakes."
Industry calls for entrepreneurial thinking
Son Henny, head of AI Education at Upstage, said companies need graduates who treat their careers with an entrepreneurial mindset. "What we desperately need are people who can suggest their own roles within a company and offer their ideas on how to navigate the AI wave together," Son said. "We need individuals who treat their lives with an entrepreneurial mindset."
Son urged universities to adopt AI across their own operations and join industry in defining new talent pipelines. "Upstage operates at the very front line, constantly probing where the boundary lies," she said. "Because these technologies are brand new, there are no established markets or ecosystems yet. We are continually defining what talent and players will be needed."
For educators looking to integrate these approaches, AI for Teachers Courses offer structured pathways for embedding AI literacy into curriculum design. The broader shift toward human-centered AI in academic settings continues to gain momentum across AI for Education initiatives worldwide.
Why this matters for educators
The panel's consensus carries immediate implications for curriculum designers, department heads, and faculty. Assessment redesign is no longer theoretical - institutions are being told directly that multiple-choice tests and standardized exams fail to develop the reasoning skills students need. The call to create environments where students can safely fail and iterate means rethinking grading policies, project structures, and how internships connect to coursework. For educators, the message is clear: the value you provide is not in delivering information AI can summarize, but in teaching students how to question, collaborate, and build on what machines produce.
Your membership also unlocks:
