Court recognizes deepfake video as subject to copyright law

The modern AI technologies enable the creation of works in science, literature, and art that can be legally considered intellectual property. However, the legal status of such objects remains a topic of discussion. A key issue is that these works are created using technological algorithms, and according to current legal regulation, a specific result of intellectual activity is recognized as a subject of copyright only if it is created by human creative labor.

According to explanations by the Supreme Court of the Russian Federation, the creative nature of creating a work does not depend on whether it was made by the author manually or with the use of technical means. However, results created by automatic technical means without human creative character are not subjects of copyright.

Thus, the essential characteristic required to recognize the result of intellectual activity as a subject of copyright is the creative nature of human activity. The result of intellectual activity is assumed to be created by creative effort until proven otherwise. It is the responsibility of the person challenging the creative nature of a work to prove otherwise.

In this regard, the decision of the Arbitration Court of Moscow dated November 30, 2023, in case number A40-200471/23, on compensation for the infringement of exclusive rights to a work – a video created using an AI system – is noteworthy.

Thus, the disputed video was created using technology based on generative adversarial networks (GAN), which allow for the imitation of reality. This effect is typically achieved by overlaying new images and video on original images or video clips. Works created using such technologies are called deepfakes.

Specifically, in the disputed deepfake video, the original image used was the visual likeness of the well-known actor Keanu Reeves, who through technical editing became the hero of a comedic situation. According to the case materials, the defendant believed the borrowed video was not a subject of copyright because it was created not by creative labor but by using deepfake technology. However, the court pointed out that the script, video shooting, and audio accompaniment of the video were created by a group of authors, and only the visual image of the actor from the original video sequence was used. Additionally, it noted that deepfake technology is not a method of creating an intellectual property object but an additional tool for editing video materials.

Based on the above, the court concluded that the use of such technology in creating the video does not mean that its authors did not contribute personal creative labor, and the work cannot be used without the copyright holder’s consent. The IP Court supported the position of the first-instance court in recognizing the deepfake video as a subject of copyright and left the decision to award compensation for its unauthorized use in force.