Meta, Microsoft y Bloomberg, demandados por herramientas de IA
Writer groups are suing major technology companies for using unauthorized copyrighted materials, specifically the controversial "Books3" dataset, to train large language models. This legal action underscores the ongoing conflict between AI developers seeking vast amounts of data to improve their models and creators demanding protection for their intellectual property, asserting that such practices constitute theft rather than innovation. The lawsuit argues that these companies derived illicit value from pirated texts without permission, challenging the industry’s reliance on “fair use” doctrines. By holding firms such as Meta, Microsoft, and Bloomberg accountable, the case seeks damages and injunctions, setting a precedent for whether current AI training methods respect copyright laws or exploit digital libraries illegally. This case is crucial for open data because it forces a reevaluation of data sourcing ethics and legal boundaries in public datasets. If successful, it could restrict the availability of copyrighted creative works in open training corpora, impacting transparency and reproducibility in AI research, or, conversely, encourage the development of more rigorous licensing frameworks for publicly accessible data.
Source: milenio.comPublished on 2023-10-20
Related news
- Universal Music demanda a Anthropic por las letras de canciones generadas por la IA
- AI Is Becoming More Powerful—but Also More Secretive
- La Cámara de Comercio de Bogotá ofrecerá una Guía de Ciberseguridad a las Mipyme
- University of Turku and SiloGen launch consortium to build the world’s largest open LLM