Copyright
What is it?
When AI generates images, texts, or music, a new and complex question arises: Who owns the created content? According to current copyright laws—in both the EU and the US—the answer is, by default: no one. AI-generated content is considered to be in the public domain because it is not created by a human. This means you do not automatically hold copyright to, for example, an AI-generated image of a dog, no matter how beautiful or unique it may be.
However, the field is beginning to shift. In 2024, the U.S. Copyright Office opened the possibility of granting rights in certain cases where a highly creative and detailed prompt expresses genuine human artistic intent. The boundaries are still unclear, and each case is assessed individually.
At the same time, there's a growing issue where AI models sometimes generate content that resembles existing works—even if they are technically "new" creations.
And behind all this lies a larger ethical dilemma: Most generative AI models are trained on content they don’t necessarily have the rights to use.
Examples:
One case involved Midjourney generating a portrait of an Afghan girl that closely resembled Steve McCurry’s iconic 1984 photo Afghan Girl. That image is protected by copyright, but the AI may have learned from thousands of similar images and ended up creating something so close that it could trigger a legal dispute. Who is responsible in such a case? The user? The provider? The AI?
In December 2023, The New York Times filed a lawsuit against OpenAI and Microsoft for using millions of their articles to train models like ChatGPT—without permission. The case isn’t just about money, but about fundamental principles of digital justice: Is it acceptable to use copyrighted content to train commercial AIs?
What to consider?
If you use AI to create content, be aware that you may not own the result—and that you could potentially infringe on someone else’s copyright if the output resembles existing work. Be especially cautious when using AI content in marketing, design, or public communication.
Ask yourself: Is it ethically defensible to use AI trained on material that was used without consent? And how can you contribute to the development of more fair and transparent models?
As legislation continues to evolve, responsible users and organizations must consider not only what is legally allowed—but also what is ethically right.