Not long ago, I was in New York talking with an investor about an AI content company. He did not begin by asking how many videos the company could produce or how quickly it could generate images. Instead, his very first question was this: “Where did the training data for that content come from? And if a rights dispute arises later, who is responsible?”
In the Age of AI Content, the Terms of Competition Are Changing
That question goes straight to the heart of the content industry in the age of AI. In the past, competitiveness in the content industry lay in who could tell the more compelling story and build the stronger fandom. But as generative AI transforms the way content is produced, the center of competition is gradually shifting.
What matters now is not only the content itself. Increasingly, what matters more is the data on which that content was built, whose creative works it learned from, the rights structure within which it is distributed, and to whom, and how, the revenue is allocated.
K-Content Has Succeeded—but Who Is Writing the Rules?
For the past two decades or so, Korea has spoken proudly of the success of K-content. K-pop, K-dramas, K-films, K-webtoons and K-food have built a powerful presence in global markets. Yet, to put it bluntly, what Korea has mainly created is content and brands. The rules governing how that content is protected, repurposed, invested in and monetized in global markets have largely been designed in the language of global platforms and foreign capital.
In the age of AI, this problem comes into sharper focus. Generative AI can serve as a megaphone for K-content, but it can just as easily become a machine for mining it without permission. The voices of Korean artists, the faces of actors, the narrative structures of dramas, the drawing styles of webtoons, and the choreography and style of K-pop can all become raw material for AI training. The real question is what comes next. When someone trains on this material and creates new content, what rights can the original creators assert? How far should platforms be held responsible? And how should we distinguish fans’ derivative works from commercial replicas produced by AI?
Consider an AI vocalist trained on an idol’s voice, a virtual advertisement using an actor’s face, generated images that imitate a webtoon artist’s drawing style, or an automated screenwriting service trained on the narrative structures of K-dramas. When these emerge, Korea must decide whether to treat them merely as infringement or to design them into a new licensing market. At this point, law ceases to be a device that holds creativity back and becomes the language of industrial design—one that ensures creators take their rightful share in new markets.
The Next Requirement for a Cultural Powerhouse: ‘K-Rules’
What matters more now than K-content is K-rules. By K-rules, I do not simply mean creating Korean-style regulation. I mean designing the legal, technological and capital structures that Korean content needs in order to keep generating value in global markets. Copyright, data-training rights, publicity rights, voice rights, portrait rights, platform liability, disclosure obligations for AI-generated content, revenue-sharing structures and licensing models all fall within this. If content is culture, rules are the means by which that culture becomes an economy.
Until now, Korea has excelled at making content. Going forward, however, it must design the structure of the market in which that content circulates. A country that only makes good content can win a fandom; a country that makes the rules takes command of the industry. American platform companies are not strong merely because their technology is superior. They are adept at turning technology into market rules. Through contracts, terms of service, data policies, algorithms, subscription models, advertising revenue-sharing and intellectual property structures, they turn technology into an industrial order.
This is why Korea cannot rest on the success of K-content alone. The fact that the name “K” has become known worldwide does not mean its value automatically returns to Korea as industrial leadership. If anything, in the age of AI that separation may accelerate. If a structure takes shape in which Korea creates the content, global platforms accumulate the data, foreign companies train the AI models, and the results are then sold back into global markets, the popularity of K may grow even as Korea’s share shrinks.
AI law, therefore, is no longer simply a matter of regulating technology. AI law is cultural policy, industrial policy and national strategy. The moment we split the question of K-content in the AI era into the Ministry of Culture, Sports and Tourism’s support programs, the Ministry of Science and ICT’s technology policy and the Ministry of Justice’s after-the-fact dispute resolution, Korea becomes a country responding belatedly within rules already designed by the platforms. Data is a question of technology and also of rights; content is a question of culture and also of capital; platforms are a question of distribution and also of national competitiveness.
In my new book, “K Must Die for K to Live,” I wrote that K can no longer remain an emotional slogan or a brand. The title is provocative, but the concern behind it is simple. The point is not to reject K, but to carry K to the next stage. In the age of AI, that concern becomes all the more urgent. If K is to survive, it must move beyond content and become the rules.
Korea must now ask: Who makes K-content, where is its data accumulated, how are its rights protected, and how is its revenue distributed? And above all, who makes the rules? If, in the age of AI, K remains only a name attached to songs and dramas, Korea will receive the applause while the money and the data pile up elsewhere. If Korea is to become a true cultural powerhouse, it must now go beyond K-content and build the K-rules on which content runs.