The rapid advancement of artificial intelligence (AI) technology has intensified debates over potential copyright infringements in AI training processes. While AI companies argue that utilizing vast datasets is essential to maintain their competitive edge, creators are demanding copyright protection for unauthorized uses and are filing lawsuits. As the boundaries between AI and copyright law are tested, governments worldwide are actively engaging in legal and policy discussions to establish new regulations.
OpenAI asserts that using copyrighted materials to train AI models is unavoidable and should be recognized as fair use. This argument is based on similar logic to Google’s book digitization project in 2005, where a U.S. court ruled that digitizing books for search and learning purposes fell under fair use. However, as the scope of copyright-protected content expands, legal issues concerning AI model training have become more complex, raising controversies regarding OpenAI’s stance.
OpenAI is currently facing multiple copyright infringement lawsuits. In December 2023, The New York Times filed a lawsuit against OpenAI and Microsoft, accusing them of using their articles without permission to train AI models. Major Canadian news outlets have taken similar legal actions against OpenAI, and digital news corporations in India owned by large enterprises are preparing lawsuits for the same reasons.
Additionally, in July 2023, writers Sarah Silverman, Christopher Golden, and Richard Kadrey filed lawsuits claiming that OpenAI had used their works without authorization as training data. These lawsuits signal a fundamental reevaluation of how AI companies collect and utilize content.
In July 2024, the U.S. Copyright Office released its first report on AI copyright issues, followed by a second report on January 19, 2025. The office examined the possibility of copyright protection for AI-generated works, emphasizing that “human creativity is the core element of copyright protection.” According to the report, content created by humans using AI as a tool can be protected by copyright, but content independently generated by AI cannot. The office highlighted the need for clear legal standards regarding the use of copyrighted materials in AI training processes.
Meanwhile, the European Union and the UK are also discussing regulations related to AI copyright. Governments are attempting to balance legal and policy adjustments between tech industries and creators while preparing guidelines for the ongoing growth of the AI industry.
The clash between AI technology development and copyright protection is evolving beyond legal disputes into a phase of exploring new collaborative models where AI and creators can coexist. AI companies are considering strategies such as entering licensing agreements with publishers and content creators or establishing fair compensation systems.
However, finding a balance between copyright protection and AI technology innovation remains a challenge. In an era where AI transcends the boundaries of creation and legal protection, attention is focused on how technology and law will harmonize and advance together.