← All news
Kokoro
Kokoro is an open-source, lightweight text-to-speech model that can generate natural-sounding speech from text. Its small size and permissive licensing make it easy for developers to embed voice output in products and internal tools, which means companies may be deploying it without a formal review. For governance teams, that raises questions about voice cloning risks, consent for any voice data, and whether speech features fall under disclosure rules for synthetic media.
1 item
