There seems to often be an idea that the models "contain" the training images. Aren't they far too small for that? A point of having a model (LLM, whatever) is that the model is abstracted from the training data. And then, except coincidence that doesn't really matter, the model cannot regenerate exactly a person's source image.
It can generate a likeness and that will be enough for people to complain about; and likeness is already actionable in narrow circumstances in current law if there was some intent. But otherwise, likeness, even very close likeness is safe in current law.
One current problem is that copyright covers "derived" works. And it will have to be sorted out whether model training means everything that follows is "derived". I sure hope for not but who knows. Music law certainly goes pretty extreme on "derived". While elsewhere we are actually pretty free for our inspiration.
The question of tools that are capable of generating illegal content is pretty safe currently. Not illegal. But there's no doubt that some large copyright owners will go after that sooner or later. It's tradition.
I definitely understand the model does not contain the images.
You gave a very valid description for probably the most critical copyright issue around it at the moment.
I'd say my point is more around the fact that the content will have different legal standing depending on whether or not something produced by the model is considered a traded good or a private creative work using some tools.
Copyright and likeness rights is only one area this impacts, but is a solid example. If the content is considered traded, then likeness is not allowed. If it is a private work created through use of a tool, then it is fine as long as it isn't traded or publicly displayed.
My argument is that since the model is heavily influenced by its training, and the user did not train the model, that a trade is happening. The untrained model is the tool, but a trained model is now like a commodity and everything it produces is also a traded item.
It can generate a likeness and that will be enough for people to complain about; and likeness is already actionable in narrow circumstances in current law if there was some intent. But otherwise, likeness, even very close likeness is safe in current law.
One current problem is that copyright covers "derived" works. And it will have to be sorted out whether model training means everything that follows is "derived". I sure hope for not but who knows. Music law certainly goes pretty extreme on "derived". While elsewhere we are actually pretty free for our inspiration.
The question of tools that are capable of generating illegal content is pretty safe currently. Not illegal. But there's no doubt that some large copyright owners will go after that sooner or later. It's tradition.