AI and Intellectual Property: Redefining the 'Human Creator' in Copyright and Patent Law

Generative AI models, particularly ChatGPT, DALL-E, and Midjourney, can produce original texts, images, music, and even software code in seconds. While this radically changes the processes of creativity and innovation, it also calls into question a centuries-old fundamental principle of intellectual property law: must the owner of a work or invention be human?

Traditional intellectual property law, in the context of both copyrights (authorship) and industrial property rights (patents, trademarks, designs), centers on the "human" factor. For example, in Turkey, the Law on Intellectual and Artistic Works (FSEK) No. 5846 defines the author as the "natural person who creates the work." Similarly, under the Industrial Property Law (SMK) No. 6769, the person who makes an invention, the "inventor," must also be a human. Legal systems are built on the assumption that creativity and inventive activity are products of human mental and intellectual effort. So, where does this traditional definition stand when an AI algorithm creates an original painting with minimal human intervention or designs a new chemical compound?

The answer to this question varies depending on the role of AI in the creation process. In the first scenario, AI is used as a "tool." Just as a photographer edits a photo using Photoshop, an artist can create a work by giving specific commands (prompts) to AI, selecting the color palette, and repeatedly editing the output. In this case, the control of the creative process and the final decisions belong to the human. Legal doctrine and current practices largely agree that in this scenario, the author is the human who uses AI as a tool. Since the creative will and original touch come from the human, it is the human who will benefit from copyright protection.

The real legal debate ignites in the second scenario, where AI ceases to be a "tool" and becomes an autonomous "creator." Imagine an AI composing an original symphony or discovering the molecular structure of a drug by learning on its own and analyzing datasets, without a specific human command beyond a general purpose. In this case, who "creates the work" or "makes the invention"? The AI itself? The programmer who coded it? The owners of the dataset that trained it? Or the company that runs and invested in the AI? The current legal frameworks cannot provide a clear answer to these questions.

Significant cases and debates are taking place on this issue internationally. One of the most well-known examples is the patent applications by Dr. Stephen Thaler, who listed his AI system named "DABUS" as the inventor. Many authorities, such as the US, the UK, and the European Patent Office (EPO), rejected these applications on the grounds that the inventor must be a human. Similarly, the U.S. Copyright Office decided in a comic book case featuring AI-generated images that only the human-authored text and the creative arrangement of elements were protected, while the images themselves could not benefit from copyright protection. These decisions show how strong the "human creator" principle still is.

So, how can the law adapt to this technological development? Three main potential scenarios are being considered. The first is that works and inventions without a human creator are considered public domain. While this approach facilitates access to information, it may be weak in encouraging the multi-billion dollar investments in AI development. The second scenario is the creation of a "sui generis"—a unique—new category of rights for AI-generated content. This could provide a more limited and shorter-term protection than copyright, protecting the investor while not disregarding the public interest. The third and seemingly most pragmatic scenario is to attribute ownership to the natural or legal person who develops, trains, or legally operates the AI. This approach can be likened to existing principles of "corporate authorship" or the "work-for-hire" doctrine.

In conclusion, the legal status of AI-generated content is one of the most current and complex issues in the field of intellectual property. Legal regulations worldwide, including Turkish law, have not yet been updated to fully respond to this technological revolution. In the current situation, creators and companies producing content with AI can create a favorable position for themselves in potential legal disputes by documenting the extent of human intervention and creative contribution to the process. In this new and dynamic field, seeking legal counsel to protect your rights and determine the right strategy is more critical than ever.