Voice translations preserve speaker's tone, pacing, pitch—with SynthID watermarks for security.

Source: Ars Technica — AI — read the full report at the original publisher.

This is a curated wire item. The Continuum Brief does not republish full third-party articles; this entry links to the original source.