Did you know Meta Claims ‘Fair Use’ in Lawsuit Over AI’s Use of Copyrighted Books

 

Meta Platforms hopes a legal decision taken by the American Federal Court goes in its favor. This has to do with a pending decision related to the company using copyrighted books to train its AI models.

Meta says it did nothing wrong, even though it failed to take consent before using content belonging to authors like Sarah Silverman and Ta-Nehisi Coates. The tech giant further went on to argue about how such activities fall under the ‘fair use’ domain. This is the name reserved for a legal document that enables limited use of copyright material without taking permission under certain situations.

During Monday’s court hearing, the company’s legal team shared with a judge how the case was filed by authors, but it had no grounds. Therefore, Meta felt it needed to be dismissed. Meta also mentioned how the Llama AI model was curated to help in a wide number of tasks, such as creative ideas, business report production, and analysis of data. This includes using books transformatively.  

Facebook’s parent firm then argued that the whole training process doesn’t aim to copy the original work belonging to authors but assists Llama in serving different practical functions. Common examples here include providing customized tutoring and assistance during writing.

The legal war also includes authors’ claims that Meta is utilizing pirated variants of their material for Llama training without getting the right kind of authorization. The company’s reply about fair use being a fundamental part of the development for transformational GenAI models is critical. It argued how it gave rise to innovative and creative processes.

Meta’s spokesperson mentioned how fair use is pivotal to transforming GenAI LLMs that are open source and power huge innovation, creative processes, and productive tasks. It also highlighted how the company keeps fair use as its top priority when utilizing LLMs.

The authors mentioned during the legal case from 2023 how Meta was engaging in infringement by literally feeding their books into the models without providing any form of compensation or taking consent.

Now, they want the court to reject the company’s defense, while asserting how this organization keeps exploiting material for expressive works that stays guarded against copyright laws. The case is being heard in the court of California right now and is certainly a high-profile one for obvious reasons.

The authors have hired a top legal team comprising several attorneys to present their case, while Meta’s defense team also includes top lawyers like Kannon Shanmugam and Bobby Ghajar.