Now, it’s returned from the dead, with regular news coverage and articles by professional-sounding journalists who have introductory notes and contact numbers against their names.
He’s one of the journalists, nonfiction writers and newspaper publishers actively campaigning against the “unauthorised” use of GenAI, which, they say, poses an existential threat to their livelihoods. “The question is whether AI systems can copy copyrighted material without compensation and for their own profit. We do not think the US copyright law allows it,” Justin A. Nelson, partner at law firm Susman Godfrey LLP, tells TRT World.
“Just as humans are permitted to learn from the ideas, concepts, and patterns in copyrighted materials, copyright law has for decades recognised that fair use principles allow intermediate copying and use of copyrighted materials for the purpose of learning and creating new, transformative works,” itHe says the argument that using copyrighted works for training chatbots is itself copyright infringement – irrespective of any similarity with the generated output – is wrong as a matter of both law...
In those instances, he says, there’s a “very strong case” that the generated output infringes copyright. But he warns that even if the output is indeed infringing, it does not automatically mean every actor involved in every stage of the training should be held responsible for it.The courts are indeed getting inundated with cases by content producers against GenAI firms.
Acknowledging that GenAI will have “significant disruptive effects” on some white-collar and creative workers, Sag says the likes of The New York Times will be “just fine”.