Una imagen de IA a partir de texto no tiene autoría humana, según USA
The US Copyright Office recently clarified that images generated solely from text prompts do not qualify for human authorship, treating these inputs more like instructions for a commissioned artist rather than creative expression. This stance establishes a crucial precedent, distinguishing between mechanical reproduction by AI and original human conception, thereby challenging the traditional notion of sole human ownership in the age of generative technologies. However, the guidelines leave room for human creators who significantly modify AI-generated material. If an individual adds substantial, visible changes that reflect their own mental conception, the resulting work may receive copyright protection. This nuance suggests that while raw AI output is unprotected, human intervention can bridge the gap, provided the artistic contribution is distinct enough to be recognized as an original work. This evolution is vital for open data and digital rights, as it forces a redefinition of creativity and ownership in hybrid workflows. It impacts how open datasets and AI tools are used legally, highlighting the need for transparent frameworks that acknowledge both technological generation and human curation. As rules are written in real-time, this case sets a benchmark for balancing innovation with intellectual property rights in emerging digital landscapes.
Source: es.gizmodo.comPublished on 2023-03-18