News Companies Sue OpenAI for Signing Deals

0:00Avoiding your unfinished home projects because you're not sure where to start? Thumbtack knows homes, so you don't have to. Don't know the difference between matte paint finish and satin? Or what that clunking sound from your dryer is? With Thumbtack, you don't have to be a home pro. You just have to hire one. You can hire top-rated pros, see price estimates, and read reviews all on the app. Download Thumbtack today. Today we have some big news in AI, and that is the fact that five major U.S. newspapers are currently suing OpenAI for copyright infringement, and they're also suing Microsoft. I think, you know, you don't want to just sue the new startup.

0:38You want to sue the guy that's actually giving them all the money, which, of course, is Microsoft. So this is kind of an interesting deal and lawsuit. I want to break it down. We've seen similar lawsuits in the past, but all of this is coming on the backs of OpenAI making a bunch of new deals with news corporations. We have a deal with Axel Springer. They just last week signed a deal with the Financial Times. And, of course, we've had them have this whole lawsuit and a deal conclusion with the New York Times as well in the past. So in my opinion, what's really going on right here is probably that OpenAI is doing deals with a lot of news publishing companies.

1:13They're paying them out essentially for their content or for whatever. And that's totally cool. But I think because of that, then, of course, we have a bunch of other companies that are like, hey, look, we want a little piece of the action, right? They don't want just the New York Times get paid out from the lawsuits. They all want to jump on. And you pretty much see this where if there's like a piggy bank, everyone wants to smash it. the more money that OpenAI gives out to other news corporations, new ones are going to jump on. And I think because OpenAI essentially trained from the entire internet, like, yeah, you can make a case for every single news company that they were included in the training set.

1:47And something important I want to say is in these lawsuits, and we'll break down more of the specifics, but like to essentially conclude that, look, OpenAI trained off of like, let's say the blacklisted the New York Times' URL, which I don't actually think they did because I think they wanted all that data, but let's say they did because they didn't want to get in some sort of lawsuit with them, that doesn't mean that they would be completely exempt from having trained off the New York Times' data because it's super easy for anyone on the internet to copy and paste the article and repost it on their own website.

2:20Now, that's, you know, step one, and maybe the New York Times could get mad at that website or whatever, especially if it's supposed to be behind their paywall or whatever. It gets a little bit trickier though, when something that's like fairly fair use is that you go to any article on the internet and you copy like a paragraph or a couple sentences and you quote them somewhere, right? You're like, as the New York Times said in their article, this blah, blah, blah happened. And also lest you think like that there's some like wholly protected class of New York Times articles, like I've seen Bloomberg and the New York Times quote things from other people as well or other blogs or you know most recently we have this huge tech debacle where we have marcus brownley who's been reviewing a bunch of ai tech and he gives he's giving it bad reviews whatever and it's this big viral moment it's on twitter yeah every major news company is quoting direct quotes from his youtube videos and they're also you know direct quotes from his x account and people that are commenting on his x posts so So like, I don't know, it's in my opinion, it's all fair use, right?

3:20Like we have things happening and there's commentary going down. So this all, I say all this because it makes it a little bit trickier when you're doing these lawsuits and you're saying like OpenAI trained off of a New York Times snippet, because the way they actually prove that is they try to go to OpenAI and say like, hey, give me like an exact quote from this article about this thing that happened. And if OpenAI spits out like a direct quote from the New York Times article, then New York Times like haha we caught you right so anyways but like it doesn't necessarily mean it's actually from the original article it could be someone quoting it and is that even bad because like it's kind of fair use for us to grab quotes from things that are happening and and talk about them I think like obviously it would be bad if you um said what happened in this specific event and it like literally copy and pasted the whole New York Times article but it's not even doing that and I think it gets one step further sorry for all of my uh analogies here but the last thing that is important is as these ai tools are becoming more and more popular even before chat gpt came out when we had uh things like jasper ai which i was using back in like two years ago a september a couple years ago i was using this thing a ton and that was kind of essentially da vinci which was an older version of chat gpt before chat gpt was released whatever the point is people were using these ai tools to rewrite articles and today you can go stick a new york Times article and did Chad Chippity get it to rewrite it, even though the New York Times might have had exclusive data on that?

4:46And then maybe the New York Times could get mad about that. But at the same time, what's the difference between that and like me reading the New York Times article and writing my opinion piece on what they said, right? Everyone's kind of doing it. So there's a lot of gray areas. So these lawsuits are a little interesting. Let's dive into exactly what's going on here. So who are the newspapers? um they are they're eight big newspapers owned by one company i think this is what's important the same investment company which is the alden global capital um i think it's important because it's not like all these guys are friends and they all get together and sue opening eye really it's alden global capital that's suing opening eye so let's make that clear um the second thing is the new york times had a similar lawsuit and they got a bunch of they had a bunch of publishing claims Up until now, they were the only one that really took legal action against them.

5:39What's interesting is that it seems like they might still be the only independent one that took legal action versus like this is an investment group. So, you know, OpenAI, they are an investment group, so they probably want money, whatever. This is it's definitely about the money. So a lot of other newspapers, including the Financial Times, Associated Press, Axel Springer, have all made specific paid deals with OpenAI, other AI companies. where they're getting millions of dollars annually to essentially get their content included. Microsoft isn't talking about this. They're not commenting on this whole story, this whole lawsuit, whereas Opening Eye is a little bit more chatty about it.

6:15But who are these news companies that essentially are one big conglomerate? We have the New York Daily News, Chicago Tribune, Orlando Sentinel, South Florida Sun Sentinel, San Jose Mercury News, Denver Post, Orange County, Register, and St. Paul Pioneer Press. Okay, these are like local newspapers, if we're being honest, right? I guess Chicago Tribune, probably one of the bigger ones in there. But otherwise, yeah, these are like local city news, really one big investment firm. so um they're all being in they're all being represented by Rothwell fig earnest and manic which is one of two law for firms that was supporting the new york times in their lawsuit against opening on microsoft so it's kind of like we get a big investment firm that owns a bunch of these newspapers they go hire the same law firm that was you know used by the new york times to sue open ai um the lawsuit was filed in the same district as the times lawsuit which um is interesting because if the same judge is chosen to oversee both of the cases, they could actually choose to combine the two complaints.

7:19Was this done on purpose? Was this random? Obviously, this was done on purpose. They see this lawsuit from the New York Times is probably going well or has a high chance of going well. So this investment firm grabs their conglomerate of newspapers. They file a similar lawsuit in the same district, hoping the same judge gets it. If the same judge gets it, he could combine the two cases. And essentially, this investment firm would be piggybacking on the New York Times case. So you can obviously see some like pretty solid financial motives for all of this, but it's interesting. Someone that apparently was familiar with this and was reporting by Axios.

7:54So, you know, I mean, the familiar with the case, I like put a caveat on that because I hate those kind of sources when there's not an actual person's name behind it. And you could get fake stuff in news all the time on this. But in any case, apparently, according to Axios, someone familiar with the Alden subsidiaries that owns the newspapers said that the paper is right now opting to sue the two firms instead of attempting to negotiate a deal, right? So the New York Times tried to negotiate a deal with OpenAI and Microsoft for a bunch of months before their lawsuit, which OpenAI said was like a surprise because they're like, I thought we're going to negotiate a deal, but it took them a long time.

8:34Evidently what that means when they say tried to negotiate a deal and didn't work, is that like they asked for way too much money or according to open ai way too much money opening i didn't want to pay the price they're asking so they're like fine we'll sue you um evidently same thing they tried to negotiate a deal and then now comes a lawsuit because they're not getting the price they want it's going to be interesting to see if they ever get the price i think uh it's going to be a precedent that you know essentially has a lot of implications for these ai companies what they're able to do so for now um alden isn't ruling out um having more of its i think they own 60 newspapers eventually join the lawsuit they're starting with eight um right now i think what that means is they like found eight newspapers that they had like really high odds of them you know being included in the data set and they're like well we don't maybe 100 know or can prove if all 60 are but they're like they're they're like not really not adding all 60 of their newspapers to this lawsuit i guess if if it comes down to it where it's like per news site they get a certain amount of money then they're like well 60s better than 80s are multiple on that so they might just like dump them all in it'll be interesting to see what happens so how does all of this work similar to the new york times lawsuit um i think at the really center of this complaint is copyright infringement claims around the essentially opening i using their articles to train the model um the newspapers right now or this investment firm is accusing opening and microsoft of quote purloying millions of the publishers copyrighted articles without permission and without payment and this is of course to make money from chad gpt so the newspaper right now is also claiming that open ai and microsoft removed copyright management information like the journalist names and titles from the work when the information um was cited so the lawsuit also includes diluted trademark claims which allege essentially that open ai and microsoft um are they didn't have like the authorization to essentially use this um and they they also say that they use the newspaper's trademarks um in the answers on chat gpt and copilot you can kind of come like join microsoft and opening eye together in this complaint because copilot is essentially chat gpt running at their back end too right so one other thing that the newspaper has also accused both these companies of is reputational damage because of the ai's hallucinations um so essentially they're like look you're like using our info and you're like using that to give people answers but then it can also hallucinate and say the wrong thing and that's bad for us because i think this this point is kind of moot because like obviously may okay maybe i guess if i straw man this or steel man this maybe what they're saying is um according to the new york times article this this this this thing happened right so maybe chad chibiti could say that and then it had like some erroneous fact and then they're like it's gonna damage a reputation because now it's associating a brand name with that the easy solution is to just like say tell chat gpt they could just hard code it to say you're never allowed to say the new york times or the washington post or whatever um there's obviously pros and cons to that it'd make it feel like less high quality stuff i think that if they want to play hardball opening eye does that and as far as like stealing their copyright and branding obviously it's like google in my opinion it's helping their branding.

11:53It's making them relevant to cite them. The last time I went to the New York Times to read an article was, frankly, never, probably because it has a paywall, but also because, I don't know, you know, I get my news in other ways. I get it from newsletters. I get it from different places. So it seems like it's becoming less relevant. If you want to stay relevant, I think it's probably great to be cited. You know, there's a higher chance that I'm going to go read a New York Times article if I see it quoted in something, even in chat GPT, than if it never says the name. So it's kind of interesting because they're like mad they're using their data and they don't want them to use their name, whatever.

12:27Let's talk about the big picture here. I think the outcome of this lawsuit is going to be really big for how AI companies incorporate news into their content. So news publishers up until now have obviously just relied on ad revenue that has come from search traffic. It's been 20 years where we've seen this. Generative AI tools can essentially wipe that out because I'm not always going to Google. I mean, honestly, I don't. Yeah, I use perplexity. I use you dot com. I use a lot of other tools instead of Google. So Google is getting disrupted in this, but I'm getting like a concise roundup of answers.

13:01I also use like newsletters, which is kind of replaced my like AI newsletters are fantastic. Shout out. And so like I use that to get a lot of my AI news and they cite the sources, but I'm not going to click on it if they give me a really concise, which I know they just use AI tools to like to like summarize an article or whatever. But if they're just going to summarize it right there, I don't need Google to find it and I don't need to read the article to get the information. So like, yeah, between the two of those is quite disruptive. So that's been the model and it's definitely getting disrupted.

13:32They're all concerned about that and they all don't want this to happen. So it's going to be interesting to see how this plays out. Opening Eye right now has already opposed both of these lawsuits. They're fighting both of them in courts. and they say that the the new york times specifically cited how opening eyes tools regurgitated verbatim copies of the new york times so they're saying like it's literally just spitting out the same thing and they say that this is a rare bug that we're working to drive to zero like in other words chai chai bt isn't designed to copy and paste so this is probably not something that happens a ton if it does it's a rare bug they're working to drive to zero as they've said.

14:14And so, yeah, obviously everything's just going to get rewritten. In any case, it's going to be fascinating to see how this goes. If you enjoyed the podcast today, if you learned something new about these lawsuits and kind of how the landscape is going to play out in the future, I would really appreciate it if you like the video or give us a review. If you're listening on Apple or Spotify, subscribe, and I will catch you in the next video.