Adobe’s Embrace of AI Faces Legal Scrutiny Over Alleged Use of Pirated Books
Adobe has recently found itself at the center of a proposed class-action lawsuit, highlighting the legal complexities surrounding the burgeoning use of artificial intelligence within the tech industry. The lawsuit, initiated by Oregon author Elizabeth Lyon, claims that Adobe improperly utilized pirated works, including Lyon’s own writings, to train its SlimLM AI model.
The SlimLM model, designed to enhance document assistance tasks on mobile devices, was reportedly pre-trained using the SlimPajama-627B dataset, which was released by Cerebras in June 2023. Lyon alleges that her works were included without consent in a dataset manipulated from the larger RedPajama dataset, known for its controversial use in AI training. This manipulation, according to the lawsuit, implies that the SlimPajama dataset contains portions of the problematic Books3 collection, encompassing approximately 191,000 books.
This legal action aligns with a broader trend within the tech community, where several companies have faced scrutiny for potentially infringing copyrights as they leap into AI advancements. Notably, both Apple and Salesforce have encountered similar allegations involving the RedPajama dataset in the past months. These legal battles not only reflect the challenges of navigating intellectual property rights but also highlight the pressing need for clear guidelines regarding the ethical training of AI systems.
Key points to consider:
– The lawsuit claims Adobe’s SlimLM model utilized pirated texts for training.
– Lyon asserts her copyrighted works were included in unlicensed datasets.
– The case raises ongoing concerns about the use of large datasets that may include protected materials in AI development.
– Similar lawsuits have surfaced recently, indicating a growing trend in legal challenges within the tech industry.
As the conversation around AI and copyright evolves, the outcome of this lawsuit could set important precedents for the future of AI training practices.
