The Rise of AI-Generated Video: A Double-Edged Sword for South Korea
The deluge of torrential rain that recently hit South Korea brought with it not only floodwaters but also a surge of hyperrealistic AI-generated videos depicting fantastical scenarios, blurring the lines between reality and fabrication. One viral video showcased Seoul’s iconic Gyeongbokgung Palace submerged in water, with individuals in yellow raincoats bailing out water, and a seal swimming across the flooded courtyard. While visually stunning, the scene was entirely fabricated, a product of the rapidly advancing capabilities of artificial intelligence.
The proliferation of these AI-generated videos is largely attributed to the accessibility of tools like Google’s Veo3, a text-to-video AI platform that empowers everyday users to create high-resolution videos with voice commands. Since its launch in May, Veo3 has fueled the creation of over 40 million videos globally, averaging 600,000 new clips daily. This democratization of AI video creation has ushered in a new era of content creation, transforming industries from broadcasting to advertising.
South Korean broadcasting giant MBC has already embraced this technology, utilizing AI to recreate historical events like the theft of the Mona Lisa and the first spacewalk for its show “Surprise.” While lauded as innovative, this move has also sparked concerns about the potential displacement of human actors, crews, and post-production staff. The ease with which AI can now generate realistic video content raises questions about the future of these professions and the potential economic implications.
Beyond job displacement, the spread of AI-generated video poses a significant threat of misinformation. Several broadcasters have inadvertently aired fabricated footage as real news, highlighting the difficulty in distinguishing between authentic and AI-generated content. One such incident involved a video of a sparrow attacking invasive bugs, later revealed to be AI-generated. Similarly, a fabricated image of an environmental activist insulting “lovebugs” during a protest went viral before being debunked. These incidents underscore the urgent need for mechanisms to identify and flag AI-generated content to prevent the spread of misinformation.
The darker side of this technology is also emerging, with a dramatic increase in deepfake-related crimes in South Korea. Cases involving romance scams and voice phishing have skyrocketed, with police reports linked to deepfakes increasing more than sixfold from 156 cases in 2021 to 964 in 2022, with no signs of abating. The sophistication of these deepfakes makes them increasingly difficult to detect, posing a serious challenge to law enforcement and highlighting the potential for malicious use of AI-generated video.
South Korea is grappling with the challenge of regulating this rapidly evolving technology. The country’s AI Basic Law, slated to take effect in January 2026, mandates watermarking AI-generated content. However, critics argue that watermarks can be easily removed or ignored, rendering them ineffective. Alternative solutions, such as embedding identification technology directly into AI models, have been proposed as a more robust approach. However, there are also concerns that overregulation could stifle innovation in the domestic AI industry. The challenge lies in striking a balance between fostering technological advancement and mitigating the potential risks associated with AI-generated content. As South Korea continues to embrace the creative potential of AI, it must confront the urgent need for responsible development and implementation to prevent the erosion of public trust and maintain social cohesion in an era where the line between fact and fiction becomes increasingly blurred.
