In its submission to the Australian governmentâs review of the regulatory framework around AI, Google said that copyright law should be altered to allow for generative AI systems to scrape the internet.
I said ârecognisableâ, and it is clearly recognisable as Gettyâs watermark, by virtue of the fact that many people, not only I, recognise it as such. You said that the models donât use any ârecognizable part of the original material that it was trained onâ, and that is clearly false because people do recognise parts of the original material. You canât argue away other peopleâs ability to recognise the parts of the original works that they recognise.
I said that models donât contain any recognizable part of the original material. They might be able to produce recognizable versions of parts of the original material, as weâre seeing here. Thatâs an important distinction. The model itself does not âcontainâ the images from the training set. It only contains concepts about those images, and concepts are not something that can be copyrighted.
If you want to claim copyright violations over specific output images, sure, thatâs valid. If I were to hit on exactly the right set of prompts and pseudorandom seed values to get a model to spit out an image that was a dead ringer for a copyrighted work and I was to distribute copies of that resulting image, thatâs copyright violation. But the model itself is not a copyright violation. No more than an artist is inherently violating copyright because he could potentially pick up his paint brush and produce a copy of an existing work that heâs previously seen.
In any event, as I said, Getty isnât suing over the copyright to their watermark.
