Publishers' Alliance Challenges AI's Fair Use Doctrine at Meta

Five major book publishers have launched a class-action lawsuit against Meta, alleging its Llama models were trained on illegally copied books. This legal challenge transcends a simple copyright dispute, representing a coordinated attack by the high-value publishing establishment on the foundational assumption of "fair use" for AI training data. By bringing in educational and professional publishers like McGraw Hill and Elsevier, this suit strategically shifts the battlefield from fiction, as seen in The New York Times vs. OpenAI case, to the lucrative domain of proprietary academic and instructional content, aiming to dismantle AI’s unregulated data acquisition model. The lawsuit’s core strategy—alleging direct "word-for-word" copying—is designed to sidestep the ambiguous fair use debate and present a clearer case of infringement, placing the burden of proof on Meta to demonstrate its training data