Adobe is facing a proposed class-action lawsuit alleging misuse of authors' works in the training of its SlimLM AI model. The lawsuit, filed by author Elizabeth Lyon, claims Adobe used copyrighted books without permission, attribution, or compensation. The core of the dispute revolves around the dataset SlimPajama-627B, which is used to train SlimLM and is allegedly derived from RedPajama, a dataset known to include the 'Books3' database containing a large number of pirated books.
Lyon's lawsuit highlights that her non-fiction writing guides were included in the training data. This legal challenge joins a series of similar copyright-related complaints against tech companies involved in AI development, including suits against Meta, Apple and Salesforce. The outcome of this case, along with others, could significantly impact the legal standards surrounding AI training data and copyright law.
