AI Literacy Is Not Just a Skill—It’s the Backbone of Responsible AI Governance, Experts Warn

Artificial intelligence is transforming nearly every corner of modern life, from healthcare and education to finance and criminal justice, but a new study from South Korea argues that the most powerful tool for managing its risks is not another regulation—it is a fundamentally reimagined form of AI literacy. The research, published in Frontiers in Communication, positions AI literacy not merely as an individual competency but as a critical component of responsible AI governance, a shift with profound implications for how nations approach the technology’s rapid diffusion.

The study, conducted by Sunkyung Shin of Woosuk University and Jooyeun Park of Hankuk University of Foreign Studies, drew on written structured expert interviews with eleven media experts from academia, industry, and national policy research institutes in South Korea. Through qualitative content analysis, the researchers sought to understand how these experts perceive the societal impacts of AI and what policy directions they believe could mitigate associated risks. The findings reveal a landscape of interconnected concerns and opportunities spanning three domains: individual, societal, and industrial.

At the individual level, all eleven participants identified enhanced convenience as a major benefit of AI, largely attributed to recommendation systems, smart home technologies, and automation services that simplify daily decision-making. Experts also pointed to personalized services, improved access to information, and support for creative activities as individual-level advantages. Yet these benefits came with significant caveats. Six participants expressed concern about personal data protection and security, noting that AI systems must continuously learn from users’ daily data, creating persistent vulnerabilities to privacy infringement and information leaks. Others worried about declining cognitive and critical thinking abilities due to excessive reliance on AI, algorithmic bias, and the deepening of the digital divide.

The societal-level analysis revealed even starker tensions. Every single expert identified increased unemployment due to AI-induced job displacement as a major societal risk. Participants described generative AI as likely to reduce demand for human labor, leading to the disappearance of certain jobs or shifts into low-wage roles. As AI moves into professional sectors, overall labor demand may decline, reducing access to quality employment and intensifying industry competition and profit disparities. Four participants also identified the potential for AI to intensify social polarization and inequality, particularly through the concentration of AI technologies and economic benefits within a small number of large firms.

At the industrial level, the experts focused heavily on the media and entertainment sector, where AI is reshaping content production across creation, editing, distribution, and consumption. Ten participants identified increased efficiency in content production as a positive industrial impact, noting that AI-based video production and personalized content delivery contribute to content quality, competitiveness, and accessibility. However, five participants also identified job reduction and role replacement as industrial-level risks, alongside growing polarization between large and small industry actors, intellectual property and copyright disputes, and the spread of fake information and news.

When asked about policy responses to these risks, the experts converged on a striking consensus: AI literacy education was mentioned by all eleven participants as a critical policy response. This unanimous agreement underscores the experts’ belief that expanding public understanding of data practices, misinformation, and human–AI interaction is essential for risk mitigation. Other policy responses included strengthening corporate accountability through self-regulation, mentioned by nine participants, and improving government regulations and legal frameworks, mentioned by seven participants. Experts emphasized that firms developing AI systems should anticipate potential harms and take proactive measures to address them, as neglecting societal impacts could trigger more stringent state intervention.

The study’s most significant contribution lies in its repositioning of AI literacy from a narrow technical skill set to a multidimensional public capability. The analysis identified four core values that should guide AI literacy policy: ethics and responsibility, practical and real-world applicability, social inclusion and accessibility, and autonomy-oriented learning approaches. Experts argued that AI literacy should encompass the capacity to recognize and manage risks such as misinformation, intellectual property conflicts, and potential societal harms, while also providing learning opportunities grounded in practical contexts and ensuring equitable access across different age, income, and social backgrounds.

The emphasis on social inclusion reflects concerns about technological polarization and unequal access to AI-related opportunities. Participants stressed the importance of ensuring that people of all ages and socioeconomic backgrounds have equitable access to AI education and resources, which they view as crucial for narrowing the digital divide and enabling everyone to benefit from AI advancements. The autonomy-oriented dimension highlights the role of user agency, critical thinking, and the capacity for independent evaluation of AI systems, with experts noting that purely regulation-centered models may be limited because coercive or top-down approaches are unlikely to be effective.

South Korea offers a particularly compelling context for this research. The country became the second nation to implement a comprehensive AI legal framework, with the Basic Act on the Promotion of Artificial Intelligence and the Establishment of a Foundation for Trust taking effect in January 2026, modeled after the European Union’s Artificial Intelligence Act. This legislative milestone, combined with vigorous investment by globally competitive firms such as Samsung, Naver, and Kakao, and a technologically literate user base, positions South Korea as a valuable case for examining how policy frameworks and literacy strategies are formulated in a rapidly digitizing environment.

The study’s findings suggest that effective AI governance requires moving beyond the traditional binary of innovation versus control. Instead, the researchers propose a multilayered governance model that simultaneously cultivates user competence, encourages responsible industry behavior, and situates regulatory authority within a broader ecosystem of shared responsibility among government, industry, and civil society. This approach reflects a shift toward co-regulation, in which multiple stakeholders jointly contribute to the development of norms and safeguards for technological advancement and social protection. The researchers acknowledge that their framework is expert-informed and interpretive rather than empirically validated, and they call for future research to engage citizens, digitally marginalized populations, affected communities, and policymakers in its participatory refinement.

Subject of Research: AI governance

Article Title: Developing an AI literacy policy framework for responsible AI governance: perspectives from South Korean media experts

Article References: Shin, S., & Park, J. (2026). Developing an AI literacy policy framework for responsible AI governance: perspectives from South Korean media experts. Frontiers in Communication, 11, Article 1966278. https://doi.org/10.3389/fcomm.2026.1966278

Image Credits: AI Generated

DOI: 10.3389/fcomm.2026.1966278

Keywords: AI literacy, AI governance, South Korea, AI policy, media experts, responsible AI, digital divide, Developing, literacy, policy, framework, responsible