Adobe is facing a proposed class-action lawsuit alleging the company used pirated books to train one of its artificial intelligence models. The complaint, filed on behalf of Oregon-based author Elizabeth Lyon and first reported by Reuters, claims Adobe relied on unauthorized copies of copyrighted works to train SlimLM, a small language model designed for document assistance on mobile devices.
According to the filing, SlimLM was pre-trained using SlimPajama-627B, an open-source dataset released by Cerebrasin 2023. The lawsuit argues that SlimPajama is a derivative of the RedPajama dataset and includes material from Books3, a controversial corpus of approximately 191,000 copyrighted books. Lyon alleges her own works were included without consent or compensation.
The case adds to growing legal scrutiny across the AI industry over training data provenance and copyright compliance.
