How AI Companies Can Pay Fair Rates for the Content They Need

Frontier AI models are trained on the accumulated digital output of humanity, which AI companies acquired essentially for free. This is a problem for both AI companies and content creators. For AI companies, future training will require new, high-quality human data. For creators, they face a choice between modest one-off licensing deals and copyright litigation, neither of which offers reasonable compensation. Yet AI companies already produce the two data sets required for pricing content, as a matter of course, every time a model is trained: data mixture and scaling laws capture how to divide the pie and how big it is. Further, collective management organizations (CMO), like those used in the music industry, offer a model for how to distribute payments. The technical objection that has kept both sides arguing in the dark—that data simply cannot be valued at scale—does not hold up. This framework offers both AI companies and content creators a fair path forward.

Raccontata dahbr.org