From Extractive to Reciprocal Co Creating Data with Speakers of Unwritten & Endangered Languages cop скачать в хорошем качестве

From Extractive to Reciprocal Co Creating Data with Speakers of Unwritten & Endangered Languages cop

This session explores how language data for AI can move from extractive practices to reciprocal, community-centred models, with a focus on unwritten, endangered, and minoritised languages. Drawing on real-world cases and policy analysis, the speakers examine how language data is collected, governed, and monetised—and what this means for consent, ownership, transparency, and community agency. The discussion maps a continuum of data governance models, from corporate extractivism to shared public infrastructures and community-led commons, and introduces ethical frameworks such as the Four Rs—Respect, Relevance, Reciprocity, and Responsibility. Highlighting initiatives like BOLI and global language data commons, the session argues for treating language not merely as data, but as a living knowledge system rooted in identity, culture, and collective rights. Digital Empowerment Foundation Website: https://www.defindia.org/ Facebook:   / defindia   Instagram:   / defindia   Twitter: https://twitter.com/DEFindia?s=20 Linkedin:   / verification