Insights

The Controversy Over the Public Nature of AI: 'Private Profit vs. Social Cost'

ㅣKarl Yang, Founder & Executive Director of KoSIF ㅣ

The entire world is in an uproar over Artificial Intelligence (AI). Moving far beyond mere technological progress, AI has now transformed into a massive physical infrastructure that shakes the very foundations of our communities, affecting everything from electricity, water resources, and labor markets to capital markets and educational institutions.

Perhaps we have been treating AI as a form of "magic that can be consumed infinitely."

In reality, however, while giant tech corporations amass staggering wealth through AI, the social and ecological costs incurred to keep its engines running are quietly being shifted onto the shoulders of ordinary citizens.

      The Reality of "Physical Exploitation" and Digital Enclosure

The AI industry has currently entered a stage of "physical exploitation." The data centers required to sustain Large Language Models (LLMs) swallow up the planet’s energy like a black hole. The International Energy Agency (IEA) warned that by 2026, the electricity consumption of data centers will exceed 1,000 TWh—an amount equivalent to the total electricity consumption of Japan. Furthermore, the massive volume of water poured in to cool down the heat generated during computation threatens the water security of local communities. AI appears to exist in a virtual space, but in truth, it drives carbon emissions and resource depletion.

The critical point we must focus on here is the ownership of "data”, the core fuel that runs this massive machine. Today’s AI has grown by learning from the countless records and intellectual assets generated by citizens in their daily lives.

Yet, the resulting power of intelligence and the financial returns are monopolized by a handful of Big Tech companies. This is nothing less than a "digital enclosure"—extracting data, which is a shared resource belonging to all of us, and converting it into private profit.

Ultimately, an irrational structure of "privatizing profits and socializing costs" is freezing into place. While corporations generate wealth using citizens' data, they pass the external costs—such as power grid congestion and environmental destruction—onto the society at large.

      Reclaiming Data Sovereignty and the Public Good

We must look at AI once again through the lens of data sovereignty. Data is not the exclusive property of corporations; it is a right belonging to citizens, and the value generated from it must be funneled back into the public interest. Leaving AI entirely to the whims of a free market compromises the very sovereignty that underpins democracy.

Therefore, treating AI from the perspective of a "public good" is an urgent historical task that can no longer be delayed. This does not mean grinding technological innovation to a halt. Rather, it means holding corporations that utilize data and resources accountable for their social impact.

Just like the carbon emissions trading system, we need to clearly define a global cap on the maximum resources used for AI. Concurrently, a portion of the profits derived from data utilization should be reclaimed—in the form of a "data dividend" or a public fund—and used as financial capital for a just transition.

      Setting Democratic Priorities

Furthermore, we must establish a "democratic priority list" for the utilization of AI resources. Top priority for resource allocation should be granted to domains directly tied to human survival, such as solving the climate crisis or public healthcare.

Conversely, strict brakes must be applied to resource consumption in areas with low public value, such as indiscriminate targeted advertising or algorithmic labor control. This means pacing technological advancement based on "community dignity" and "data sovereignty," rather than market efficiency alone.

The climate crisis has left us with a stern lesson: infinite growth with unpaid costs eventually triggers the collapse of the entire system. The true threat of AI lies not in intelligence itself, but in the logic that legitimizes the exploitation of public resources like data and the environment. Only when we reach a collective consensus on whom it is used for and how far it should go can technology truly become progress for everyone.