SOCAN is testing whether AI attribution can support payment

SOCAN has entered an exploratory relationship with Canadian rights-technology company Musical AI. The pilot will test whether attribution tools can identify when existing music influenced or was used in AI-generated output.

The first stage uses a mutually agreed selection of publicly available generated material and Musical AI’s existing technology. SOCAN says any non-public information used in the test will be deleted afterwards, and member data will require explicit permission.

This is a test, not a new royalty system

Attribution is one missing technical layer in AI licensing: creators cannot claim payment for a use that nobody can reliably identify. But recognition alone does not set a price, establish authorization or distribute money.

The practical question is therefore whether the system is accurate enough to become evidence inside a rights workflow. If it is, collecting societies would still need agreements that define which uses count, how contributions are valued and who pays. SOCAN has started evaluating the measurement layer, not announced that those policy questions are solved.