◆이코노미스트가 글로벌 차세대 경제 리더, 청소년 기자단 영 저널리스트와 함께합니다. 영 저널리스트 기자단은 프리미엄 경제지 이코노미스트, 논술 전문 기관 Ni 에듀케이션과 함께 주요 시사 이슈를 팔로우업하고 직접 기획, 취재, 기사 작성 활동을 하며 사회적 문제를 고심하고 해결 방안을 제시하는 과정을 경험하게 됩니다. 이번 기사는 영 저널리스트 기자단이 현 사회현상에 대해 학생들 시선에서 ‘왜’라는 질문을 갖고 직접 취재해 작성한 기획기사입니다. 영 저널리스트 기자단의 기획기사는 영문과 국문, 두 형태로 게재합니다.  

[Young journalist Cheon Ye On] Currently, AI technology is climbing up to its apex. From AlphaGo to ChatGPT, AI is showcasing enormous growth in the short term. For instance, Suno AI can create catchy music in a few seconds, and OpenAI’s ChatGPT not only writes text, but also creates artworks that seem to be made by real artists. 

Moreover, as the trend of ChatGPT converting original photos into Ghibli style became popular, AI users have increased globally. Finally, the time for countless people to freely express their ideas, minds, and creativity has come.

The AI industry doesn’t always exhibit its bright side. AI is trained by a process called Deep Learning. Deep Learning, a method of AI training, collects data from images and writings on the internet. Throughout this process, AI ‘learns’ various knowledge, utilizing it in order to reply appropriately to the users and create artworks for them. AI acquires knowledge about  human-made creation, and mixes these numerous data, and creates the following result. This process automatically makes AI create similar works from other human creators.

As these situations occur, this problem may be a natural event to happen: The problem centers on AI and its copyright issue.  People are debating whether AI violates or doesn’t violate copyright law, creating various interpretations.

For instance, the Ghibli picture trend is a prominent example of this. Would AI generated Ghibli images infringe copyright? Most people would definitely realise that the image is referenced from Ghibli’s original style. These world-renowned companies, such as Giblish are distinguished easily as their own product. Ghibli pictures have their own unique and warm style. Therefore, there would be nobody who claims “AI plagiarized Ghibli’s image!” among those who enjoyed converting their photos into Ghibli style.

As a result, in this case ownership violation doesn’t seem to be an issue.

Nevertheless, identifying this cleanly as not plagiarizing any other human made works wouldn’t always be the right case. There has been a massive increase of companies and copyright holders that have sued AI industries such as Open AI. The creator of the global hit series ‘Game of Thrones’ sued the AI industry for using the series to train AI without approval. Various media companies such as the New York Times are famous for suing Open AI. At the end of last year, SBS, MBC, and KBS that are Korean media industries sued Naver, appealing disapproval to use the media companies’ creation without any acceptance. SBS, MBC, and KBS requested Naver to pay the compensation. As such, AI companies are being sued due to the reason that they used the original creation without any approval.

As a result, AI creation becomes a path for every person to express their own creativity. However, illegally training AI with creator’s works without their approval brings massive damage in the creators’ ownership, finance, and content company’s value.

But this doesn’t mean that the AI copyright issue isn’t unsolvable at all. Labeling the original reference of AI creation would effectively address this problem. Whenever the user is publishing AI-generated images and other types of works, the user must label that this creation is AI-generated, and if the AI creation is referenced from original human made works, the user should credit the exact reference of the creation. Korea already is about to implement ‘AI Basic Act’, a  mandatory law of inserting watermarks in any AI generated videos.

Concurrently, copyright protection law should be enforced. AI companies should pay ‘royalties’ to the creator, earning their right to train the creator’s work for Deep Learning. This method would protect creators’ ownership. According to IT columnist Han Sae-Hee, it is said that “recently media industries such as  AP and Axios have signed a contract to provide contents to AI company, and world’s most popular internet community, Reddit monetized AI companies’ data training” and that “Earning profit from AI’s content learning is the way of AI and the creators coexisting”, about paying copyright-fee to the creators.

Through protection of artworks, writing, and other types of creation, we should lead rapidly improving AI technologies in a safe and beneficial path. This is the time to try various methods in order to obstruct AI’s danger.

ⓒ이코노미스트(https://economist.co.kr) '내일을 위한 경제뉴스 이코노미스트' 무단 전재 및 재배포 금지