Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
An exception to this UMG v. Anthropic In this case, because at least early on, previous versions of Anthropic generated lyrics for the songs in the output. This is the problem. The current state of this case is that they’ve put in a lien to try to prevent that from happening, and the parties have kind of agreed that those liens are sufficient pending the resolution of the case, so they’re no longer looking. preliminary instruction.
At the end of the day, it’s no longer a difficult question for AI companies Is it legal to practice? this What do you do when your AI comes up with a very similar result to a particular case?
Do you expect most of these cases to go to trial, or do you see settlements on the horizon?
There may be some settlements. Where I expect to see settlements are big players with great content or especially valuable content. The New York Times may end up with an agreement and license agreement in which OpenAI pays to use New York Times content.
There’s enough money at stake that we’ll probably get at least some judgment that sets the parameters. The class plaintiffs, in my opinion, have stars in their eyes. There are many class actions, and my guess is that the defendants will resist them and hope to prevail in a final judgment. It is not obvious that they will go to court. In the Supreme Court Google vs. Oracle He pushed very hard for the right of fair use to be settled by summary judgment rather than before a jury. I think the AI companies will work hard to get these cases resolved on a common decision basis.
Why would it be better for them to win a summary judgment against a jury verdict?
It’s faster and cheaper than testing. And AI companies worry that they won’t be considered popular, which many people will think. Oh, you copied something that should be illegal and do not delve into the details of the fair use doctrine.
There have been many deals between AI companies media outletscontent providers and other rights holders. Most of the time these deals seem to be more about the search than the main models, or at least that’s how it was described to me. In your opinion, should licensing content be used in AI search engines – the answers are sourced with search augmented generation or RAG – something that is legally binding? Why do they do that?
If you’re using search-expanded generation on targeted, specific content, your fair use argument becomes more difficult. An AI-generated search is more likely to produce text taken from a specific source in direct access, and less likely to be fair use. That is, he can maybe, but the risky area is that it is more likely to compete with the original source material. If, instead of directing people to a New York Times story, I give them my AI suggestion that uses RAG to extract text from that New York Times story, that seems like a tradeoff that could hurt the New York Times. The legal risk for an AI company is greater.
What would you like people to know about generative AI copyright battles that they may not already know or may be misinformed about?
What I often hear mistaken as a technical matter is the notion that these are simply plagiarism machines. All they do is take my stuff and send it back in the form of texts and replies. I hear a lot of artists say that, and I hear a lot of lay people say that, and it’s technically incorrect. You can decide whether generative AI is good or bad. You can decide whether it is legal or illegal. But this is really something new that we haven’t experienced before. Having to practice a bunch of content to understand how sentences work, how arguments work, and understand different facts about the world doesn’t mean just copying and pasting things or making a collage. It really creates things that no one expected or predicted and gives us a lot of new content. I think it is important and valuable.
