Humans vs. machines: the fight to copyright AI art

A new high-stakes argument has erupted over the authorship of AI-generated art. Last year, artist Kris Kashtanova fed a new AI program his instructions for a graphic novel, which produced stunning artwork based on his prompts. The resulting story, titled “Zarya of the Dawn,” was about a future Manhattan with a character resembling actress Zendaya. While Kashtanova initially received a copyright for the story, the U.S. Copyright Office abruptly revoked this in February, claiming that the images were “not the product of human authorship.”

This has led to an ongoing debate over who holds the rights to AI-generated artwork. People are curious to know if the owners of the AI systems or the users could secure copyright protection. Billions of dollars could ride on the decision, as companies would use AI to generate low-cost images, video, and text for branding, advertising, and entertainment if they could hold the rights. However, the Copyright Office has long held that the author of a work must be a human being, and many artists and owners who fiercely hold art rights object to granting copyrights to AI owners or users. Lawsuits for copyright infringement cases have been filed by owners of AI programs such as chatGPT, Stability AI, and Midjourney, which Getty Images also separately filed against.

Other owners claim that the new algorithms use copyrighted material from the open web, rendering AI-based systems illegal due to copyright infringement laws. While several people have argued that their AI programs are sentient and should be recognized as creators of the artwork, so far, the courts have continually rejected the argument.

Meanwhile, Kashtanova is testing the limits of the law with a new book, using the Stable Diffusion AI program, which allows users to start with original artwork and refine them with text prompts. This time, Kashtanova hopes, there will be enough of a “human” element for the authorities to sway their side. Kashtanova plans on submitting copyright applications for individual images from the new autobiographical comic, providing a series of methods, settings, and AI programs.

The legal system is yet to figure out the answer to the AI-generated artwork authorship issue, and the new technology appears to pose a significant challenge to the traditional legal system.