Can Artificial Intelligence Violate Copyright? Three Real Cases That Spark Debate

Can AI violate copyright? Explore three current legal cases and the ethical challenges of copyright law in the age of artificial intelligence.

Who truly owns a work created by an algorithm? If an image, a song, or a text is generated by a machine, can we speak of copyright? Artificial intelligence is redefining the boundaries of creativity, but also those of intellectual property. In recent months, numerous cases have raised far-reaching legal and ethical questions. Let's look at three that have sparked significant debate.

The Case of Zarya of the Dawn: AI Art Put to the Test by the Copyright Office

In 2022, Kris Kashtanova submitted a comic book titled Zarya of the Dawn, generated with the help of Midjourney, to the U.S. Copyright Office. The registration was initially accepted but later revoked: according to the Copyright Office, images created by AI cannot be copyrighted because they lack "human authorship." Kashtanova pointed out that the composition, script, and direction were entirely hers, but that was not enough. The case opened an international debate: where does human work end and algorithmic work begin?

Stability AI and Getty Images: AI and Datasets Under Scrutiny

Another example involves Stability AI, the company behind the Stable Diffusion model. Getty Images has sued it, alleging the use of millions of copyrighted images to train the algorithm. The issue touches on a fundamental point: is it permissible to use protected material to train an AI, even if the result is something "new"? Currently, case law is uncertain. But the Getty vs. Stability lawsuit risks setting a precedent for the entire industry.

This case also connects to the article already published on La Bussola dell’IA, "AI and Copyright" where we delved into the legal and technological dilemmas related to the topic.

The AI Musician: Drake's Voice Without Drake

In 2023, a song apparently performed by Drake and The Weeknd went viral on TikTok. In reality, it had been generated by an artificial intelligence. The song was removed from Spotify and other platforms at the request of the record label, but by then millions of people had already listened to it. Who holds the rights to a track created by an AI but with the "voice" of a real artist? Is it imitation, satire, or infringement?

The case has also raised questions regarding the protection of vocal identity, an aspect still poorly regulated in current legal systems. In Europe, the debate is just beginning but is moving towards forms of protection similar to the right to one's image.

Between Law, Ethics, and Innovation: A Gray Area to Explore

These episodes show how the law is struggling to keep pace with the speed of innovation. Copyright, born in an analog era, struggles to adapt to hybrid works, where the hand of man and that of the machine intertwine. It is no coincidence that the European Union, in the proposed AI Act, is beginning to discuss responsibility and transparency in AI-generated content【useful external link: Proposal for an AI Act Regulation – EUR-Lex】.

Ethics, for its part, raises questions about the value of human creativity. If a machine can imitate the style of Van Gogh, Chopin, or Stephen King, what remains of artistic uniqueness? And most importantly: who should profit from this content?

On La Bussola dell’IA we have already covered the relationship between technology and imagination in the article "AI and Creativity", which we recommend reading for a broader perspective.

An Open Challenge

Artificial intelligence is redefining the rules of the creative game. Copyright is one of the most fragile areas in this transition, but also one of the most fascinating. The three cases discussed show us that there is no single answer yet. But one thing is certain: the line between innovation and abuse must be drawn carefully, and preferably, in advance.