Paul McCartney says a sound editor used AI to “extricate” John Lennon's voice from an old demo to create “the final Beatles record”, slated for release in 2023
Sir Paul McCartney says he has employed artificial intelligence to help create what he calls “the final Beatles record”.
Context & Ripple Effects
This is an early, high-profile use of AI as an archival-production tool rather than a standalone synthetic performance. It follows the documented use of AI to recreate Anthony Bourdain’s voice for a documentary, making the Beatles project a prominent test of how audiences distinguish restoration from re-creation.
Later coverage shows the project became the released “Now and Then” recording, and its subsequent Grammy recognition validated the finished track’s industry reception. That arc gives this initial disclosure significance beyond a one-off studio technique.
First-order effects
- AI processing makes a low-fidelity Lennon vocal source usable for McCartney and the remaining Beatles production team, supporting the planned final-record release.
- The project immediately ties the Beatles’ legacy catalog to a disclosed AI-assisted workflow, rather than presenting the resulting vocal as newly recorded material.
Second-order effects
- A landmark artist project gives labels, producers, and platforms a concrete example of AI-assisted archival restoration as voice tools move toward commercial music use; later reporting described YouTube pursuing label rights for tools using famous musicians’ voices as platforms explored artist-voice AI.
- The distinction between extracting an existing performance and generating a new one becomes commercially important for provenance, credits, and how audiences assess posthumous vocal work.
Third-order effects
- If similar projects proliferate, valuable catalogs could become an ongoing source of AI-assisted releases, pushing rights holders to formalize approval and disclosure practices for archival voices.
- The broader market may separate restoration-oriented AI from synthetic voice products, though that boundary will depend on how artists, estates, labels, and platforms apply it.
The trend: AI is moving from experimental voice synthesis into the commercialization and restoration of established music catalogs.