1 min read

OpenAI’s ChatGPT training may be a ‘problem’ for some authors

OpenAI, the Microsoft-backed artificial intelligence research startup, is facing a lawsuit filed by two US authors in San Francisco federal court. The authors, Paul Tremblay and Mona Awad, claim that OpenAI’s generative AI tool, ChatGPT, has infringed upon their copyrights by using data copied from thousands of books without permission.

ChatGPT, which launched in November 2022, quickly gained popularity and reached 100 million active users within two months. However, the authors allege that the data used to train the AI tool was sourced unlawfully from nearly 300,000 books, including illegal “shadow libraries” that offer copyrighted works without permission.

Tremblay and Awad argue that their writings have been incorporated into ChatGPT’s database, as the AI system can generate accurate summaries of their books. The lawsuit represents a proposed class action on behalf of other copyright owners in the US who claim that OpenAI misused their works.

The lawsuit seeks unspecified damages from OpenAI, with other copyright owners also filing legal challenges against the use of their data to train AI systems. Some plaintiffs have targeted OpenAI and Microsoft’s GitHub, while visual artists have accused other image-creating tools. The companies being sued have defended their use of copyrighted data, stating that it was done fairly.

Leave a Reply