For years there has been a war on the internet between the holders of intellectual property rights, be they record companies, film producers, publishers or the media, and the technological giants. And the contest reaches a new level with the universe created by artificial intelligence generative (IAG). Intelligences that, in order to function, need to be trained with billions of contents from the network, texts, images, sounds, so that they make their statistical inferences and give those answers that seem prodigious.
Are tech companies taking advantage of the intellectual property rights of others? Without using those works with protected rights for their training, as they can do in many countries, would they have the same capacity? Does training an artificial intelligence with rights-based texts or images mean that its responses will then derive from them? Is a generative artificial intelligence a cut and paste or can it resemble Picasso entering the Prado, contemplating thousands of works and then creating?
The answers are not always simple, but the swords are high. And authors and publishers, visual artists and the media ask that it be necessary to request their authorization to use their contents and that the IAG products clarify that they are and what their training databases have been. EVA, the association of European Visual Artists, has launched the IA Manifesto, in which it recalls that there is no “artificial intelligence art without human works of art, and that they cannot use it without authorization”.
And there is special concern in the media not only because these new platforms can train their systems with their news and use it to respond to their users, but because they see a serious threat to democracy: the results they offer can be fed from unreliable sources but not are subject to no liability for veracity.
Lawsuits have started to break out in the US. The large Getty photography database has sued the imaging platform Stable Diffusion, claiming that it used 12 million of its images to train its AI model without permission or compensation. And comedian Sarah Silverman has sued Open AI (ChatGPT) and Meta, Mark Zuckerberg’s platform, for training, she says, his language models with his books.
Andrés Guadamuz, professor of Intellectual Property Law at the University of Sussex, believes that the lawsuit against Meta is the one that may have the longest run. “OpenAI points out that it used 85% of web content, Wikipedia, Reddit, many databases, and 15% books for training the model. One part without rights and another that is not clear. But Meta did use a database protected by copyright: 169,000 books, ”he points out.
And remember that in many countries there are exceptions for those uses created to advance AI research. The EU, after the 2019 directive on copyright in the digital single market, allows it, he points out, “for scientific institutions, but also for companies like Meta, as long as they respect the opt-out -the right to request the exclusion – of the beneficiaries”. The same happens in Japan or the United Kingdom, “but in the US it is still not very clear.” Although, he points out, the case is even more complicated “because the database used by Meta was not created by them but by a non-profit organization, EleutherAI.”
“I think that Meta’s defense -he warns- is going to be that in the database used, each one of the books does not have independent commercial value. They could try to reach a collective agreement with the publishing houses to pay for a license. Another issue raised by the lawsuit is that the model is a derivative of the books used in training. I don’t think so, and I think no judge. It’s 169,000, you remove Sarah’s books and the model works the same. What matters is having thousands and thousands of works, billions of words, and looking for statistical patterns in them. I don’t think we could legally talk about a derivative.”
In Europe, he says, with the current legislation, lawsuits have a difficult path. “A photographer has sued the non-profit German image database Laion, but I don’t see a future for it.” In fact, Laion has sent the photographer a demand for 887 euros for improper claim. But Guadamuz does acknowledge that “artists have a better economic case to go against the databases. There is no doubt that many illustrators are going to be out of a job. They can ask for existing exceptions to be changed. And the same with the musicians. AI companies have not wanted to release music products for fear of the industry, but they already have great models,” she notes.
Jorge Corrales, general director of Cedro, an entity that manages the reproduction rights of books, magazines and newspapers, is clear that in this field “any use of protected content must request authorization from the authors or publishers. And we need to know what contents are used in the product that they are serving to the market”. And he points out that the current European directive is already obsolete. “Brussels is working for a new framework. There is a draft text that requires that the content generated by the machines carry an identification that it has been generated with AI. And that it provides a summary of the contents used”.
From the Information Media Association, its general director, Irene Lanzaco, asks to respect “the right of information publishers to endorse the work effort undertaken by them and that all use by generative artificial intelligence is subject to authorization from the publisher and to the payment of adequate economic compensation”. “And generative artificial intelligence has to answer for the reliability of its information with the same scope with which a press editor answers. We cannot continue to pretend that technology platforms have no responsibility for the content they disseminate, because they are in a position of absolute primacy. There can be no benefit without responsibility. And it would be good to call on Europe to protect the value of its cultural creation. In the very essence of Europe is the cultural production that this continent has offered to the world. It is our main asset and we cannot trivialize it in this revolution”, she concludes.
