A growing wave of writers, illustrators and musicians is taking artificial intelligence companies to court over the unauthorised use of their work to train AI models, with several lawsuits progressing and some artists claiming early wins. The Verge profiles a number of these cases, including that of author Kirk Wallace Johnson, who discovered his books had been pirated and used to train a chatbot after The Atlantic published a searchable dataset of training material. The dispute matters because it is testing, in real time, how copyright law applies to generative AI, with implications for how much protection creators will have as the technology becomes more embedded in everyday use.

Cases cited include a class action brought since January 2023 by cartoonist Sarah Andersen, Karla Ortiz, Kelly McKernan and other visual artists against Stability AI, Midjourney, DeviantArt and Runway AI, as well as separate legal action against Anthropic led by law firm Susman Godfrey on behalf of authors. Other artists have since launched similar claims against Meta, Google and AI music generator Suno, mostly on copyright grounds though some also cite breaches of terms of service. Outcomes have been mixed, with wins, losses and lengthy delays largely hinging on how courts interpret "fair use," but those involved remain broadly optimistic that the litigation will help shape clearer legal boundaries for AI development.