Voice Data Governance Workshop

​This session is part of the Introduction to Political Technology course at Newspeak House, open to faculty and fellowship candidates only.

​Who owns your voice? And more specifically, the language you speak?

​The problem: small language communities are poorly represented on the internet. Voice assistants do not support smaller languages, neither do most websites or translation tools.

​To remedy this, various open source initiatives allow communities to upload their spoken language, which can then serve as training data for AI systems. However, while such open source systems enable the survival of small languages, they potentially also allow for their exploitation of small languages by bigger platforms.

​In this interactive session (yes, there will be role playing) we'll explore the different complications of making language data openly available. We'll also delve into different governance models and safeguards that might help communities retain control.