Content gatekeeping in a GenAI world
The training data fight is over. The access fight is just starting.
For two years now the conversation about AI and content has been stuck on one question: who gets paid for what was already used to train the models? It is a real question with billion-dollar answers, but it is also a question about the past. Meanwhile a different question has been quietly assembling itself in the background — who gets paid when a model goes to fetch something right now — and almost nobody is talking about it, even though the plumbing is already getting installed.
