New Nonprofit Launches to ‘Certify’ Copyright-Friendly AI Practices
Fairly Trained introduces a certification for generative AI companies, distinguishing those that obtain explicit creator consent for training data from those relying on legal gray areas like fair use. This initiative highlights a growing industry divide, aiming to identify and promote models that treat creators fairly by securing proper licensing rather than ignoring copyright obligations. The launch responds to rising litigation and shifting industry norms, as major tech firms face lawsuits over unauthorized data use while media giants negotiate licensing deals. By verifying that models avoid unlicensed copyrighted works, the certification provides a clear metric for ethical AI development, encouraging a market shift toward transparency and mutual respect between technology providers and content creators. This development is crucial for open data because it establishes a new standard for ethical data sourcing in the AI ecosystem. It underscores the importance of consent and licensing in data practices, urging the open data community to prioritize rights management. As AI models increasingly rely on vast datasets, this certification offers a framework for balancing innovation with creator rights, influencing how future open data initiatives handle copyright and attribution in generative technologies.
Source: publishersweekly.comPublished on 2024-01-18
Related news
- This Tech Exec Quit His Job to Fight Generative AI's Original Sin
- Mario Mendoza se refirió a la inteligencia artificial y abrió debate sobre derechos de autor - Pulzo
- La economía de China registra uno de sus peores crecimientos en décadas
- UK water industry’s ‘urgent’ plan to tackle sewage pollution delayed by four months