Can AI Replace Human Empathy? The Warning Over Welfare’s Future

Asan Foundation Chairman Chung Mong-joon delivers remarks at the foundation’s 49th anniversary academic symposium./Asan Foundation

As artificial intelligence (AI) rapidly expands into welfare administration and counseling settings, academics warn that human dignity and accountability must take precedence over efficiency. Experts said that as AI use grows, welfare systems should establish human-centered values and clear ethical standards alongside technological adoption.

On the 28th, the Asan Social Welfare Foundation said it hosted an academic symposium titled \”Social Welfare Innovation in the AI Era\” at the Asan Institute for Policy Studies in Jongno-gu, Seoul.

Marking its 49th anniversary this year, the Asan Foundation has held an annual symposium on major social welfare issues since 1979. At this year’s event, the foundation presented leading research results from its 2025 academic research support program.

In his opening remarks, Asan Foundation Chairman Chung Mong-joon said that as AI use expands in the social welfare sector, empathy, relationship-building and ethical judgment among people will become more important. He added that even while deploying advanced AI, institutions must not lose sight of human dignity.

Academic experts debate with presenters at the Asan Foundation’s 49th anniversary symposium./Asan Foundation

Hwang Kwang-seon, a professor of public administration at Gachon University, presented on the ethical challenges and limits of AI-based social welfare. He emphasized that AI should support human judgment rather than replace it, and that reducing administrative burden cannot justify abdicating moral responsibility.

Kim Hyun-jung, a professor at Dong-A University’s Graduate School of International Studies, reviewed welfare data-integration cases from Estonia and the U.K. She argued that South Korea needs a hybrid AI welfare model that combines Estonia’s legal enforcement mechanisms with the U.K.’s phased management approach.

Oh Young-sam, a professor of social welfare at Pukyong National University, said AI shows strong potential for document and organizational management but is limited in direct service delivery. He argued the social welfare sector requires selective and differentiated approaches where appropriate.

Experts specifically called for discussion about how far human oversight can and should extend when AI systems begin performing welfare checks and counseling for older adults living alone, particularly regarding decision-making standards and accountability.

Oh Seon-jung, a professor of public administration at Chonnam National University, warned that AI adoption could spread faster than expected and urged proactive, preemptive measures grounded in critical perspective.

Jin Hyo-jin, a professor of computer engineering at Gyeongsang National University, stressed the importance of empirical, data-driven verification to understand what decisions and changes AI actually produces in real-world practice.