Holly+ reframes voice cloning as consent-first
Scientific American profiles Holly Herndon’s Holly+ project, a real-time AI voice model that lets others sing in her voice under explicit licensing terms. The story positions it as an artist-led alternative to unconsented deepfake voice cloning.
This is less a new product drop and more a policy experiment for generative media: control, consent, and revenue-sharing built into the tool itself.
- –Holly+ has existed since 2021, so the news angle is renewed mainstream attention, not a fresh launch.
- –The project shows a practical middle path between banning voice AI and letting unlicensed cloning run wild.
- –For AI builders, the key takeaway is product design around permissions and provenance, not just model quality.
- –Community reactions center on legal enforceability, which remains the weak point even when consent frameworks exist.
DISCOVERED
83d ago
2026-03-05
PUBLISHED
85d ago
2026-03-03
RELEVANCE
AUTHOR
scientificamerican