Why K-Content Stars Say AI Still Needs a Human Heart

A Seoul entertainment forum has turned one of K-content's biggest technology questions into a human one: if artificial intelligence can imitate faces, voices and production styles, what still makes a performance feel alive? At the 2026 K Forum held July 9 at Conrad Seoul in Yeouido, 2PM member and actor Hwang Chansung and film director Lee Jae-kyu argued that AI can expand Korean entertainment, but it cannot replace the human heat that makes audiences care.

The session, framed around the idea of K-star intellectual property in the AI era, brought together two useful perspectives. Hwang spoke as a performer whose image and career could now become data. Lee spoke as a director watching AI move from a novelty into production planning. Their discussion captured a turning point for Korean entertainment: the industry is no longer debating whether AI will enter the creative process, but how to use it without weakening the trust between performers, creators and fans.

The timing matters because Korean companies are already experimenting with AI-centered content. Related coverage around the forum highlighted EO Contents Group's plan for what it describes as Korea's first seasonal AI drama format, with short episodes, AI human performance and expandable story worlds. That made the forum feel less like a distant theory session and more like a preview of the next production battlefield.

Hwang Chansung's Case for Human Presence

Hwang began from a place K-pop fans understand immediately: live performance. He referred to 2PM's recent Tokyo Dome concert marking the group's 15th anniversary in Japan and said the overseas response felt dramatically different from the group's earlier years abroad. His point was not simply that K-pop has grown bigger. It was that Korean culture now has a visible everyday presence, from concert venues to overseas stores where Korean products occupy recognizable space.

That observation gave his AI comments weight. Korean entertainment became global not because it was efficient, but because fans attached themselves to particular voices, gestures, mistakes, endurance and emotional histories. A star's value is not only the image on screen. It is the knowledge that a real person stood on a stage, carried a scene, took a risk and built memory with viewers over time.

Hwang did not reject technology. He said early encounters with AI advertising felt fresh, especially the possibility of completing a commercial without physically filming every frame. But he also described a concern shared by many actors: if a performer's face or body can be reproduced through digital data, then rights, consent and usage rules must become much clearer. Without that framework, AI can turn a star's identity into a resource others can exploit.

That distinction is crucial in K-entertainment. A performer is often the emotional center of a whole business ecosystem. Fandoms buy albums, watch dramas, attend concerts and follow interviews because they believe in the person behind the content. If an actor's likeness or an idol's image appears without transparent permission, the issue is not only legal. It threatens the emotional contract that fandom depends on.

Hwang's strongest point was about the energy of actual work. He argued that the heat produced when performers sweat through a stage or actors struggle through a scene is captured on screen in ways audiences can feel. The result may be technically imperfect, but that very unpredictability can make it alive. In his view, AI may imitate the surface of performance, but it cannot fully take away the force created by a human being in a specific moment.

Lee Jae-kyu Sees AI as a Tool With Limits

Director Lee Jae-kyu widened the discussion from celebrity likeness to production structure. He suggested that once a performer's digital data is available, many forms of content could be created around it within the next few years. For producers, that possibility is obvious: digital doubles, alternate versions, advertising assets, localization and short-form extensions could all become faster and cheaper.

Yet Lee's most interesting point was not about speed. He contrasted AI's ability to organize patterns perfectly with the strange incompleteness of human-made work. Human stories carry contradictions, gaps and unplanned moments. For him, that imperfection is where creative warmth appears. It is also one reason Korean dramas and films often travel well: they are willing to mix sincerity, awkwardness, humor, pain and melodrama without polishing every edge away.

Lee's comments also placed AI inside the global rise of Korean storytelling. Korean series have repeatedly reached major audiences on streaming platforms, proving that Korean characters and emotions can stand at the center of worldwide pop culture. The challenge now is expansion. If AI helps smaller teams produce more ambitious work, it could lower barriers. If it encourages factories of empty content, it could dilute the very qualities that gave K-content its influence.

Hwang and Lee both pointed toward coexistence rather than refusal. AI can be useful in dangerous action scenes, accident sequences and physically demanding production environments. It can help protect actors, test visual ideas or make certain projects possible with fewer resources. The better question is not whether AI should be used, but where human agency must remain non-negotiable.

That means performers should control their digital likeness. Directors should decide why AI belongs in a scene. Producers should disclose when a performance is AI-assisted or synthetic. Audiences should not be asked to form emotional trust around something whose authorship is hidden from them.

AI Drama Plans Put the Debate to the Test

The discussion became more concrete because EO Contents Group has presented a plan for seasonal AI dramas including Soon, Night Falls and Soon, We Go to Work. The projects are described as short-form dramas made of five-minute episodes, with ten episodes per title and an emphasis on story continuity. One title is reportedly a suspense thriller set in a prison, while the broader strategy focuses on building AI-assisted story worlds that can keep expanding.

The company has described its approach as story-first AI content, using the technology as a creative partner rather than a one-off gimmick. It has also pointed to AI learning processes based on facial muscles and expression data, aiming to make AI human performances feel more natural. Those details address the central weakness many viewers notice in AI video: it can look impressive for a moment, but it often struggles with emotional continuity over time.

A seasonal AI drama is therefore a serious test. Television depends on memory, consequence and emotional accumulation. Viewers need to believe that a character carries yesterday's wound into today's choice. If AI can support that, it may become a meaningful production tool. If it cannot, the result may feel like polished fragments rather than drama.

The reported ambition is large. Related coverage mentioned a broader "Soon" series plan on the scale of 127 projects. That number shows why producers are interested. Short AI-assisted formats can be created quickly, localized efficiently and distributed to mobile audiences. But scale is not the same as loyalty. K-drama fans may sample a new format, but they stay when characters feel specific and emotionally earned.

This is where Hwang and Lee's warnings matter. If AI content is only abundant, it becomes disposable. If it helps creators reach new forms while preserving authorship and emotional truth, it could become part of the next Korean wave.

What Comes Next for K-Content

The clearest takeaway from the 2026 K Forum is that Korean entertainment is trying to define AI before AI defines it. Hwang gave performers a language for adaptation and protection. Lee gave creators a language for experimentation without surrendering the value of human imperfection. The emerging AI drama market will test whether those ideas can survive commercial pressure.

For fans, the issue is not abstract. K-pop and K-drama fandoms are built around recognition: a voice break, a glance, a rehearsal clip, a line reading, a stage habit or a behind-the-scenes moment that confirms the performer was really there. AI can copy patterns, but fandom rarely forms around pattern alone. It forms around presence.

That is why live events may become more valuable as digital content becomes easier to generate. Concerts, fan meetings, interviews, rehearsals and on-set footage provide proof of effort. They remind audiences that the person they follow is not only an image but a worker, artist and public figure making choices in real time.

At the same time, the industry cannot ignore AI. The practical future will likely be hybrid: AI for previsualization, safety, localization, secondary content and experimental formats; humans for consent, meaning, performance and final creative judgment. The winners will be companies that use AI to support emotional storytelling rather than replace it with smooth emptiness.

The forum did not answer every question about law, ethics or production practice. But it clarified the stakes. AI can make content faster and cleaner. K-content became powerful because it often feels personal, vulnerable and direct. The next phase of the Korean wave may depend on whether the industry can use the first truth without losing the second.