Open Data

¿Puede Chile liderar en IA sin modernizar su ley de propiedad intelectual? | Diario Financiero

¿Puede Chile liderar en IA sin modernizar su ley de propiedad intelectual? | Diario Financiero

The article highlights a pivotal U.S. court ruling that established that training Large Language Models on legally obtained content constitutes “fair use,” while prohibiting the use of pirated material. This legal precedent underscores that AI development is not merely a technical process but involves significant reproduction of creative works, directly impacting intellectual property rights. The decision emphasizes that innovation is valid only when it does not rely on unauthorized copies, setting a critical distinction for global AI practices. In Chile, this issue is urgent due to a growing boom in AI initiatives backed by public institutions, which currently operate under a legal framework lacking a specific exception for text and data mining. Unlike the U.S., EU, or Japan, Chilean law technically classifies the training of AI models as copyright infringement unless strictly limited to public domain materials. The recent elimination of a proposed TDM exception in the draft AI Law exacerbates this risk, leaving startups and universities vulnerable to legal sanctions and placing Chile at a competitive disadvantage in the global AI race. This situation is crucial for the open_data community because it demonstrates how rigid copyright laws can stifle the creation of necessary datasets for local AI models. Without a robust Text and Data Mining exception that balances creator protection with innovation needs, Chile risks driving AI development into informality or hindering it entirely. The article argues that updating intellectual property laws is essential to comply with international safeguards and foster a secure environment for open data utilization and AI sovereignty.

Source: df.cl
Published on 2025-11-14

Related news