AI Should Be Citizen-Owned Infrastructure, Korean Civic Tech Leader Argues

At the 2026 East Asia Civil Society Conference in Seoul, Kwon Oh-hyun, director of the social cooperative Pati, advocated for a fundamental redesign of artificial intelligence systems. He argued that AI should be owned and controlled by citizens, transforming it into a democratic infrastructure that enhances people's lives, rather than being dominated by big tech.

“

AI should be designed as a democratic infrastructure that citizens own and decide upon, enriching citizens' lives.

Kwon pointed to the dual nature of internet technology, which has expanded citizen participation but also facilitated surveillance and manipulation. He emphasized that the benefits of data accumulation, while enabling knowledge and automation, have disproportionately concentrated among a few platforms, often at the expense of workers. This duality, he explained, underscores why focusing on structural changes like ownership and control is more critical than the inherent nature of the technology itself.

He identified existing strengths in South Korean society that could foster "democratic AI." These include a history of citizen mobilization during crises, established civil society and labor union organizations, high digital accessibility, and a track record of "civic tech" initiatives where citizens have used technology to solve social problems. To realize democratic AI, Kwon stressed the importance of creating a space for citizens within AI governance, establishing digital public squares for agenda-setting, fact-checking, and dialogue, and ensuring that those affected by AI decisions participate in the process.

“

We must create a space for citizens within AI governance. This requires digital public squares where citizens can propose agendas, verify facts, and engage in dialogue and action.

Furthermore, Kwon advocated for AI as a public good, asserting that no single layer of AI, be it services, data, models, or computing infrastructure, should remain the exclusive property of a specific company. He urged a shift in perspective, viewing data not merely as raw material for industrial growth but as a collective asset accumulated from citizens' lives and labor. The development of a public mask app during the COVID-19 pandemic, a collaboration between civic developers like Pati and Code for Korea using public data and technology, was cited as a prime example of this approach.

“

No single layer of AI, service, data, model, or computing infrastructure, should remain the exclusive asset of a specific company. We must view data not just as raw material for industrial activation, but as a collective asset accumulated from citizens' lives and labor.

Ultimately, the goal is to ensure AI enriches citizens' lives. Kwon cautioned against defining the "AI basic society" simply as one where everyone proficiently uses generative AI. Instead, he envisioned a society where essential digital services in areas like labor, welfare, education, finance, culture, safety, and the environment are guaranteed as rights. This necessitates communities developing and operating their own necessary technologies and services, free from the influence of overseas big tech companies' pricing policies and system changes. He invoked the Joseon Dynasty's Silhak (practical learning) philosophy of "utilization, welfare, and virtue," advocating for an assessment of technology not only on its usefulness and productivity but also on its ability to enrich everyone's lives and who controls its benefits.

“

The AI basic society should not be defined simply as a society where everyone skillfully uses generative AI. It must be a society that guarantees necessary digital services as rights in essential life domains such as labor, welfare, education, finance, culture, safety, and the environment.