Source: Private Law Theory

Using Australia as a case study and drawing on 22 interviews with stakeholders, the author examines whether compulsory licensing and extended collective licensing (ECL) can meaningfully facilitate copyright licensing for AI development while delivering equitable remuneration to diverse right holders. Compulsory licensing could lower transaction costs by giving AI developers access to large repertoires via collective management organisations, but it risks disproportionately constraining exclusive rights and stifling emerging direct licensing markets. ECL is presented as a less intrusive alternative, yet its utility may be limited if major right holders choose to opt out of collective schemes.

Both mechanisms face significant structural limitations tied to the territorial nature of copyright law and the national treatment principle, which the author warns could divert a substantial share of remuneration to foreign right holders rather than to local creators. The article concludes that neither compulsory licensing nor ECL alone provides an optimal solution and proposes three reform principles: consider compensation at AI deployment rather than development stage; ensure remuneration collected from local AI companies reaches local right holders; and secure a fair share of payments for original creators as well as secondary right holders. These recommendations aim to align licensing design with practical distributional and market concerns identified in the interviews.