A class action lawsuit has been filed against OpenAI, the company behind the popular ChatGPT artificial intelligence (AI) program. The lawsuit alleges that OpenAI’s use of copyrighted material to train its AI models constitutes “systemic theft on a massive scale.” The plaintiffs, which include prominent authors George R.R. Martin and Jodi Picoult, argue that OpenAI’s ingestion of copyrighted works without permission infringes on their intellectual property rights. The case could have far-reaching implications for generative AI and the use of copyrighted material in training algorithms.
Generative AI programs like ChatGPT use algorithms to select words and generate responses based on lessons learned from scanning vast amounts of text from the internet. The authors argue that OpenAI’s use of copyrighted material as training data is a reproduction of their works without their consent. OpenAI is accused of benefiting from authors’ work that publishers would otherwise pay them to create. Questions remain about the exact data sets used by OpenAI and the extent to which copyrighted material is included in the training data.
OpenAI may argue that their use of copyrighted material falls under the “fair use” exception to copyright protection, which allows for limited reproduction of works for purposes like commentary or criticism. The company may claim that their use of authors’ work is transformative and distinct from simply copying the text. Similar arguments have been made by Meta, the parent company of Facebook, in response to a lawsuit filed by comedian Sarah Silverman over the use of her memoir in an AI product.
The outcome of the lawsuit could determine the future of generative AI and the use of copyrighted material in training algorithms. If the authors’ claims are upheld, AI companies may need to obtain permission from authors and publishers to use their works, potentially leading to licensing agreements. Conversely, if OpenAI prevails, it could establish a precedent allowing widespread scanning of the internet and the creation of AI models based on the acquired data. The decision could reshape the industry and have significant implications for the information marketplace.