逐字稿 🔗
WEBVTT
00:00:00.160 --> 00:00:23.519
We have economists showing us that these AI pricing algorithms can enable companies to coordinate and to raise price for consumers in a way that potentially evade any kind of detection because now the law is really the framework we use to detect collusion is really based on human conspiracy.
00:00:26.719 --> 00:00:31.359
You're listening to Bite-Size Business Law, where we discuss big issues in small doses.
00:00:31.519 --> 00:00:40.880
This podcast is sponsored by the Fordham University School of Law Corporate Law Center, which is run by Fordham Law Professor Richard Squire and me, your host Amy Martella.
00:00:41.119 --> 00:00:43.600
Okay, let's get down to business.
00:00:46.640 --> 00:00:48.960
Welcome back to Bite-Sized Business Law.
00:00:49.119 --> 00:00:52.560
This is part one of our two-part series on algorithmic pricing.
00:00:52.880 --> 00:00:54.799
Has AI been out there colluding?
00:00:54.960 --> 00:01:05.920
The conundrum today's guest confronts is whether our traditional methods of detecting collusion under antitrust law can survive in light of AI's ability to fix prices without any need to communicate at all.
00:01:06.239 --> 00:01:15.040
Giovanna Masaroto is a lecturer in law at Penn's Carrie Law School, where her scholarship focuses on how antitrust and regulatory law should adapt to emerging technologies.
00:01:15.280 --> 00:01:16.239
Giovanna, welcome.
00:01:16.640 --> 00:01:16.959
Thank you.
00:01:17.120 --> 00:01:19.120
Thank you so much, Amelia, for having me.
00:01:19.519 --> 00:01:20.400
Thank you for being here.
00:01:20.640 --> 00:01:24.640
I really enjoyed reading your paper, Rethink Inclusion in AI Algorithmic Markets.
00:01:24.799 --> 00:01:31.920
But why don't you tell us first a little bit about you and how you came to take an interest as a scholar in that intersection between tech and antitrust?
00:01:32.239 --> 00:01:32.959
Yeah, sure.
00:01:33.200 --> 00:01:48.159
So since uh early in my career, in essence, uh while I was uh a PhD student, I was working in an economic consulting firm and dealing with uh cases concerning big tech corporations, software.
00:01:48.400 --> 00:01:54.640
So the expertise of this company was uh antitrust IP, the tech sector in the tech sector.
00:01:54.879 --> 00:02:08.159
So I always felt uh required for a lawyer also to understand the underlying technology, to understand how we can enforce the law effectively in this uh different framework.
00:02:08.400 --> 00:02:17.520
So I find it fascinating to see how technologies completely change the way markets and the same businesses behave.
00:02:17.919 --> 00:02:30.479
So I started becoming increasingly interested in uh algorithms, uh, in software, and engaging with computer scientists and yeah, working in the space of AI and so on.
00:02:30.800 --> 00:02:31.360
Very cool.
00:02:31.599 --> 00:02:32.960
Why don't you tell our listeners?
00:02:33.120 --> 00:02:43.520
I mean, we're gonna take this concept for granted throughout this discussion, but why don't you start us out with what is a pricing algorithm and what types of companies that we would know actually use them?
00:02:43.919 --> 00:02:53.840
An algorithm is uh a set of instructions so that is required to have a computer to perform a specific task.
00:02:54.000 --> 00:02:58.319
So I love uh the uh analogy of Noam Chomsky.
00:02:59.199 --> 00:03:02.639
He said, uh, okay, computers are like paperweight.
00:03:02.800 --> 00:03:07.120
They do nothing without a software that tells the computer what to do.
00:03:07.280 --> 00:03:10.719
And a software is a set of algorithms.
00:03:11.120 --> 00:03:20.400
And so you can also ask a computer to set prices of uh goods and services through an algorithm.
00:03:20.719 --> 00:03:23.280
And so I assume that most companies are using algorithms.
00:03:23.360 --> 00:03:27.280
I mean, no one's sitting around the boardroom anymore going, how much should we price this t-shirt at?
00:03:27.520 --> 00:03:27.599
Right?
00:03:27.759 --> 00:03:29.840
Like Amazon's using pricing algorithms.
00:03:30.240 --> 00:03:31.039
Yeah, exactly.
00:03:31.199 --> 00:03:31.520
Yeah.
00:03:31.759 --> 00:03:40.560
Most companies, and of course, uh including Amazon, are widely using these pricing algorithms, including AI pricing algorithms.
00:03:40.719 --> 00:03:50.560
There is a variety of pricing algorithms, and of course, companies are increasingly relying on these uh algorithms to set prices of goods and services.
00:03:50.879 --> 00:04:02.479
The question you're asking is whether Section One of the Sherman Act whether it should continue to attach any weight to this evidence of communication when it's algorithms that can actually collude without needing communication.
00:04:02.639 --> 00:04:15.759
So, why don't you give our audience kind of a quick primer on Section One of the Sherman Act and how courts typically identify collusion in human conduct and specifically on how courts have evolved this concept of agreement?
00:04:16.160 --> 00:04:26.800
Section one of the Sherman Act is the main provision we have in the US to address collusion, like broadly in any context.
00:04:27.040 --> 00:04:37.279
And so, but this uh law was developed considering human conspiracy, so human to human communication and assent.
00:04:37.839 --> 00:04:55.519
However, now these uh AI pricing algorithms don't need to communicate and don't need uh yeah to shake their hands or typically think of uh conspiracy collusion as uh competitors meeting in a smoked field room.
00:04:55.680 --> 00:05:08.480
Uh but like, yeah, now AI can uh basically coordinate price among competitors without uh yeah the need to engage uh in the traditional human conspiracy.
00:05:08.720 --> 00:05:12.000
You've got your obvious that you just meant, you've got your obvious like direct evidence.
00:05:12.079 --> 00:05:15.759
You're meeting in a smoky room, competitors are saying, ooh, let's fix these prices.
00:05:15.839 -->
Load full transcript
00:05:19.680
But most of the time, courts have to infer agreement from like circumstantial evidence.
00:05:19.920 --> 00:05:26.160
So you write about this pivotal case in the 80s, Monsanto, and that's where these so-called plus factors came in.
00:05:26.240 --> 00:05:30.399
So tell us what would a court usually look to to infer that an agreement has been made?
00:05:30.560 --> 00:05:33.839
And is this actually still the standard that courts use today?
00:05:34.160 --> 00:05:40.959
Section one of the Sherman Act uh requires an agreement to infer collusion.
00:05:41.199 --> 00:05:49.199
And uh Monsanto is the test on which uh courts rely today to infer an agreement.
00:05:49.360 --> 00:05:55.680
And this uh test uh stems from Monsanto that dates back uh the 1984.
00:05:56.000 --> 00:06:02.800
So the time uh we weren't really thinking about uh AI uh pricing algorithms.
00:06:02.959 --> 00:06:16.879
So at that time, the big problem was how can courts distinguish nerve parallelism in price that can stem from independent rational conduct of competitors.
00:06:17.040 --> 00:06:21.040
So let's take the the example of gas stations.
00:06:21.199 --> 00:06:25.759
Gas stations don't need to agree to set the same price, and that's legal.
00:06:25.920 --> 00:06:30.079
So you don't want to punish companies for doing something that is legal.
00:06:30.240 --> 00:06:40.240
So Monsanto clarified: okay, it's very rare that we find explicit agreement because these agreements are typically secret because illegal.
00:06:40.480 --> 00:06:46.800
So, but I need to distinguish uh, I cannot punish mere price parallelism.
00:06:46.959 --> 00:06:55.839
So I need to find a way to infer the agreement considering the so-called paradox of proof.
00:06:56.079 --> 00:07:19.759
So the solution was uh okay, the main element uh that is required to distinguish illegal coordination from rational independent conduct is that plaintiff needs to prove to show the conscious commitment to a common scheme.
00:07:20.079 --> 00:07:35.759
So and to show this conscious commitment uh to a common scheme, courts develop these plus factors focusing on human conspiracy, so the motive to conspire.
00:07:35.920 --> 00:07:41.680
So they look for evidence of human communication, so meetings or exchange of emails.
00:07:41.920 --> 00:07:55.600
And so this is why I would say Monsanto led courts uh to focus on this circumstantial evidence rather than economic evidence that can lead to the gas station example.
00:07:55.680 --> 00:08:06.639
So, but the problem is that uh yeah, as you can imagine, this uh framework is becoming a little bit not up to the technologies we are dealing with today.
00:08:06.959 --> 00:08:08.079
So I understand what you're saying.
00:08:08.160 --> 00:08:10.319
So there's this idea of parallel conduct, right?
00:08:10.399 --> 00:08:15.360
Like every gas station in my town has the same exact price, but that's not illegal.
00:08:15.519 --> 00:08:19.279
So they need something more, and in
Monsanto, they said we need some more evidence.
00:08:19.439 --> 00:08:30.480
And you write in your paper that, you know, examples of these plus factors, like you said, emotive, communication channels between rivals, you know, anything acting contrary to your interests.
00:08:30.720 --> 00:08:35.919
Those are all like things that humans do that would stand in as evidence of collusion.
00:08:36.000 --> 00:08:42.480
But what you're saying, Giovanna, is none of these plus factors are terribly applicable to AI pricing algorithms.
00:08:42.720 --> 00:08:48.320
Yeah, so I had the opportunity to build uh an AI system years ago with the professor of AI.
00:08:48.480 --> 00:09:07.120
So I really had the sense uh that uh this AI, the AI that we are using today is basically statistics because it learns from a huge amount of data how to perform a task increasingly well because uh AI relies on thousands, millions of examples.
00:09:07.279 --> 00:09:18.559
And so the problem with AI and collusion is that I don't think it's worth uh to think okay, is AI capable of agreeing like humans?
00:09:18.960 --> 00:09:32.639
But rather we should focus on uh objective mechanisms for coordination that we know are required to maintain a supra competitive price, so price above the competitive level.
00:09:32.960 --> 00:09:56.960
And uh we have economists showing us that these AI pricing algorithms can enable companies to coordinate and to raise price for consumers in a way that potentially evade any kind of detection because now the law is really the framework we use to detect collusion, is really based on human conspiracy.
00:09:57.200 --> 00:10:09.039
So this is why AI basically automates these uh collusive strategies that economists told us are required to maintain high prices at scale.
00:10:09.360 --> 00:10:21.840
This automation that AI is capable of by learning from this huge amount of data changes the paradigm and the way we should think about collusion.
00:10:22.000 --> 00:10:29.519
Indeed, my paper titles is rethinking collusion in AI algorithmic markets.
00:10:29.759 --> 00:10:37.360
What I found especially creepy, actually, was you point out that AI algorithms can collude without communicating, right?
00:10:37.519 --> 00:10:44.320
They're not meeting in a room, they're not talking to competitors, they're just basically scraping the internet and making highly sophisticated statistical predictions.
00:10:44.480 --> 00:10:49.600
But they also collude and learn these strategies without any instructions to collude.
00:10:49.919 --> 00:10:50.159
Yep.
00:10:50.399 --> 00:10:51.279
Yeah, thanks, okay.
00:10:51.679 --> 00:10:51.919
Yes.
00:10:52.159 --> 00:11:00.320
So what you're saying, Giovanna, is that you're noticing or you're observing that AI is actually learning what you call cartel enforcement.
00:11:00.480 --> 00:11:02.559
We're talking about antitrust cartels, of course.
00:11:02.720 --> 00:11:05.840
And there's certain strategies that traditionally have gone along with that, right?
00:11:05.919 --> 00:11:08.639
Like monitoring behavior, uh punishing.
00:11:08.720 --> 00:11:17.679
Like tell us about what you've observed AI actually learning in this process that you think can be used to, like you said, rethink our approach to collusion.
00:11:17.919 --> 00:11:46.639
It's particularly important because we have algorithms where you can basically simply instruct the algorithm to fix the price, or we have AI algorithms where you don't like instruct the algorithms to fix the price, but the algorithms learn the collusive strategies because usually the instruction that uh software developers uh give to these AI algorithms is maximize profits.
00:11:46.879 --> 00:12:00.080
So and the AI algorithms learn from the market that the best way to maximize profits is by colluding with the other algorithms and set the same price.
00:12:00.399 --> 00:12:08.480
The reason why I don't propose a new law, but I propose a new evidentiary framework is exactly for this reason.
00:12:08.720 --> 00:12:18.320
So we don't know how the AI algorithms will perform in advance because it's learning from a huge amount of data.
00:12:18.480 --> 00:12:26.639
However, we can understand what is learning after all, exposed, and the antitrust enforcement is exposed.
00:12:27.120 --> 00:12:50.159
And so the the reason why I propose this uh new evidentiary framework that is based on uh objective strategies, mechanisms for coordination that we know from economists is because of this, because uh AI doesn't necessarily collude, but it can, it's likely, it can it can happen.
00:12:50.399 --> 00:13:06.159
So I think we we should distinguish the good AI for pricing and the bad AI, and we should enable companies to use AI pricing algorithms responsibly and clarifying what is legal and what is illegal.
00:13:06.240 --> 00:13:11.919
And I hope uh my my solution can illuminate from this perspective.
00:13:12.159 --> 00:13:18.799
Yeah, I think that's important to note too, because you point out in the paper that not all algorithmic pricing is harmful.
00:13:19.039 --> 00:13:21.279
Some of it is actually beneficial to consumers.
00:13:21.360 --> 00:13:23.759
So you're not saying like, hey, we got to cut this out.
00:13:23.840 --> 00:13:29.279
You're just saying this is a test to kind of snuff out the bad AI algorithmic pricing.
00:13:29.440 --> 00:13:36.240
So, what exactly is your three-pronged test for a court to apply to figure out whether there's been agreement in algorithmic pricing?
00:13:36.480 --> 00:13:43.200
As you correctly mentioned, not all AI pricing algorithms uh collude so are harmful.
00:13:43.360 --> 00:13:48.480
I think preventing all kinds of AI algorithms is not the solution.
00:13:48.720 --> 00:14:04.080
The three-pronged evidentiary framework that I propose is based on economic studies on how these algorithms can engage in collusion and uh harm consumers by raising price.
00:14:04.240 --> 00:14:18.559
So the three-pronged uh
framework that I developed in the paper consists of a first prong according to which uh plaintiffs should show the persistent supra competitive price.
00:14:18.720 --> 00:14:33.919
So, first plaintiff show me that uh the algorithms uh is uh charging supra competitive price, so price above the competitive level and for a period of time.
00:14:34.080 --> 00:14:40.799
So it's not just a one-shot high price for consumers, but it's uh persistent.
00:14:41.120 --> 00:15:02.879
The second prong is that uh what I require in the second prong is that the plaintiffs also show that this uh high price stems from the high algorithms that learn the collusive strategies that economists tell us are required to maintain uh supercompetitive price.
00:15:02.960 --> 00:15:06.159
So this high price for consumers persistently.
00:15:06.399 --> 00:15:44.080
For example, the AI algorithms is capable to monitor rivals' price and to detect rivals' deviation from the high price and to punish this deviation, so they can learn the different algorithms how to maintain the high price because they engage in a price war and they also know that it's worth to, in the short term, to have some uh profit losses in the expectation to maintain a high price in the long term.
00:15:44.480 --> 00:15:56.320
So they are very smart in the way that they understand uh market behavior and they can end up, okay, let's coordinate, because we are here to maximize price.
00:15:56.480 --> 00:16:05.519
I argue in the paper, these two prongs, if they are met, they raise a presumption of AI collusion.
00:16:05.679 --> 00:16:08.720
So the third prong is the rebuttal.
00:16:08.960 --> 00:16:30.960
So I argue that if the plaintiff meets these uh two prongs, the defendant afterwards has the opportunity to rebuttal and to show or that the price is not super competitive, or that the high price has nothing to do with these collusive strategies.
00:16:31.200 --> 00:16:44.159
So, in other words, the AI algorithms did not learn how to behave like a cartel, did not learn the strategies that economists told us that are required, like those that I just mentioned.
00:16:44.240 --> 00:16:49.200
So it's not able to punish the competitors that is lowering the price and so on.
00:16:49.440 --> 00:16:54.000
And so I think this is a good solution for multiple reasons.
00:16:54.159 --> 00:17:16.240
Also, because I think it incentivizes companies to audit their algorithms and to use these algorithms responsibly, because they know that if the high pricing algorithms end up raising the price persistently and learns these strategies, the algorithms violates section one.
00:17:16.720 --> 00:17:25.680
So let's say a court applies this framework, and at the end of a given lawsuit, the court finds that a harmful pricing algorithm actually has been used.
00:17:25.839 --> 00:17:27.119
What's the remedy?
00:17:27.279 --> 00:17:28.880
What will the court order there?
00:17:29.200 --> 00:17:40.319
Yeah, that's a wonderful question because uh one of the issues that led to Monsanto was uh the challenge uh for courts to crafting remedies uh in the paradox of proof.
00:17:40.400 --> 00:17:51.599
So in the situation of uh uh when uh, in the example of gas station, you cannot uh distinguish between uh in illegal coordination and independent rational conduct.
00:17:51.759 --> 00:17:55.680
So remedies, how to craft remedies is uh critical.
00:17:55.920 --> 00:18:09.200
If uh courts use my proposal, my framework, my evidentiary framework, they can understand how the algorithms learn the collusive strategies and behave like a cartel.
00:18:09.279 --> 00:18:22.799
So we should be capable to understand if it was a problem of data, of confidential data, or a specific feature of the algorithms that led the algorithms to collude.
00:18:23.200 --> 00:18:41.519
So I would say we can target the remedy case by case in a way that I believe is more effective, because what we are seeing is that companies are very fast in adapting their technology to the law to enable to evade the law.
00:18:41.759 --> 00:18:46.720
A good general remedy, I think, that could be a compliance program.
00:18:47.279 --> 00:19:00.720
In a sense, uh, you company have to submit to audit your algorithms and submit the report to the DOJ or the FTC to make sure that your algorithm is not behaving like a cartel.
00:19:01.200 --> 00:19:04.960
And in fact, courts are grappling with some of these issues right now.
00:19:05.119 --> 00:19:10.880
So, Giovanna, tell us about the circuit split, the current circuit split that's arisen in AI algorithmic pricing.
00:19:11.119 --> 00:19:24.480
As you mentioned, courts uh are dealing with this issue constantly, in a sense, it's uh it's a not topic before courts right now, because as mentioned at the beginning, companies are increasingly using these AI pricing algorithms.
00:19:24.799 --> 00:19:40.559
And interesting enough, yeah, we have a decision from the ninth circuit uh uh related to AI pricing algorithms in the casino hotel room, and the use of these AI pricing algorithms by casino hotels.
00:19:41.039 --> 00:19:49.119
And in this case, the ninth circuit, so the Court of Appeal uh basically confirmed the dismissal of the case.
00:19:49.200 --> 00:19:52.400
So it was identified the section one violation.
00:19:52.720 --> 00:20:16.400
Conversely, uh, before the third circuit last July in Caesar, in the Caesar case, the third circuit reversed the dismissal in favor of the defendant in a case that is almost identical because it concerns the use of a high pricing algorithm in casino hotels.
00:20:16.720 --> 00:20:29.599
And the algorithms that these casino hotels used is almost identical to the algorithms that the other casino hotels used in the case before the ninth circuit.
00:20:29.920 --> 00:20:37.599
Caesar, which is uh one of the defendants in the case before the third circuit, is also defendant in the other case.
00:20:37.839 --> 00:20:48.240
There are people that argue this is
not real a real split because in the 9th circuit case it was dropped, they have been spoke claim.
00:20:48.400 --> 00:20:52.319
However, I think this case is an emblematic example.
00:20:52.480 --> 00:21:12.000
Like these two cases represent emblematic examples of how we need a consistent solution to address the problem because we have another case that was settled with the consent degree between the DOJ and uh Real Page, one of the leading algorithmic companies.
00:21:12.240 --> 00:21:22.480
So I think this battleground that we are seeing in AI collusion is uh requiring a consistent way to address the problem.
00:21:22.799 --> 00:21:26.079
Yeah, and of course the Supreme Court no longer cares about circuit splits.
00:21:26.240 --> 00:21:28.160
So we will see what happens with that.
00:21:28.400 --> 00:21:33.200
But as you mentioned, states are kind of taking up the mantle here and addressing the issue.
00:21:33.359 --> 00:21:41.519
Now, here in New York, we have a really interesting case going on because we have a New York law, right, that prohibits landlords from using algorithmic pricing to fix rent.
00:21:41.759 --> 00:21:52.480
It was recently struck down by a Southern District, federal judge in the Southern District of New York on a preliminary injunction motion, but it was based on this like free speech rights of the landlords argument.
00:21:52.640 --> 00:21:53.440
What's going on there?
00:21:53.519 --> 00:21:54.720
That's really kind of an interesting case.
00:21:54.799 --> 00:21:56.079
I wonder if you have any thoughts about that.
00:21:56.400 --> 00:21:56.880
Yeah, sure.
00:21:57.039 --> 00:21:59.119
Yeah, I'm not a First Amendment scholar.
00:21:59.279 --> 00:22:02.480
However, a Found the case fascinating.
00:22:02.720 --> 00:22:04.880
The case was brought by RealPage.
00:22:05.119 --> 00:22:14.960
So the company, the AI software company that I just mentioned, that is involved in the landmark case that was uh brought by the DOJ.
00:22:15.440 --> 00:22:28.160
In this case, uh, RealPage argued okay, this law is uh preventing my speech because it's basically preventing my communication of price recommendations.
00:22:28.319 --> 00:22:31.359
And price recommendations is not illegal.
00:22:31.759 --> 00:22:38.319
So this is where like RealPage uh found uh the First Amendment uh violation.
00:22:38.559 --> 00:23:00.880
And uh yeah, it's I found honestly interest very interesting the opinion, and uh basically the judge uh granted a preliminary injunction against this law, and it would be interesting to see the reactions of other states, California as uh a state legislation that aims to address algorithmic collusion as well.
00:23:01.200 --> 00:23:10.960
So, Giovanna, what is your sense of the appetite that the Trump administration has for dealing with or reining in AI algorithmic pricing?
00:23:11.440 --> 00:23:16.240
The Trump administration uh honestly showed interest in the matter.
00:23:16.319 --> 00:23:39.039
In a sense, uh there was
uh a speech last May by the uh Deputy Assistant Attorney General for Criminal Enforcement at the DOJ Antitrust Division, where we see that yeah, person from the DOJ argue that software doesn't change the rule, so competitors uh must compete.
00:23:39.200 --> 00:23:44.559
The collusion uh is not changed by the fact that companies are using algorithms.
00:23:44.720 --> 00:23:55.759
And so I, from what I read, uh the the Trump administration seems interested uh in uh addressing AI algorithmic collusion.
00:23:56.079 --> 00:24:02.880
In the landmark case that I mentioned also before, real page, there was a consent decree.
00:24:03.279 --> 00:24:13.279
Consent decrees are widely used in antitrust because uh antitrust cases are very expensive, uh, can take forever, can take a long time.
00:24:13.359 --> 00:24:15.279
Uh, we have seen with the Google case.
00:24:15.359 --> 00:24:22.319
So and these markets change so fast that it's really hard also for a judge to decide.
00:24:22.720 --> 00:24:26.079
So it's uh it's common that cases are settled.
00:24:26.160 --> 00:24:32.559
And this administration uh showed a particular interest in setting antitrust cases.
00:24:32.799 --> 00:24:39.599
Yeah, it seems like the government's first instinct is always to say, oh, uh, the frameworks we already have in place are fine.
00:24:39.759 --> 00:24:40.720
They don't need to change.
00:24:40.799 --> 00:24:46.240
We can apply our typical rules to crypto, we can apply our typical rules to AI.
00:24:46.480 --> 00:24:53.680
And that's why it's so important that we have people like you out here kind of pointing out the nuance and saying, not really.
00:24:54.000 --> 00:24:59.839
AI is different from humans, and you need to use these objective factors rather than these plus factors.
00:24:59.920 --> 00:25:03.119
And so I'm grateful that you're out here writing about these issues.
00:25:03.200 --> 00:25:07.039
It's so super interesting, Giovanna, and we thank you so much for your time.
00:25:07.200 --> 00:25:19.359
Like I said, our next episode will be part two of this algorithmic pricing series, where we'll talk beyond antitrust with Mathuz Grachowski, and we'll talk about other legal frameworks that might exist to rein in algorithmic pricing.
00:25:19.440 --> 00:25:21.359
But Giovanna, thank you so much for your time today.
00:25:21.440 --> 00:25:22.240
This was fabulous.
00:25:22.559 --> 00:25:22.799
Thank you.
00:25:22.960 --> 00:25:24.240
Thank you so much, Amalia.
00:25:24.400 --> 00:25:27.759
This has been really fun, and I enjoy it immensely.
00:25:27.839 --> 00:25:28.400
So thank you.
00:25:28.559 --> 00:25:28.960
Thank you.
00:25:29.119 --> 00:25:29.920
Thank you.
00:25:32.720 --> 00:25:38.000
Thanks to everyone at Fordham Law School, especially the Dean's Office and our corporate law center donors for supporting this podcast.
00:25:38.240 --> 00:25:41.119
Bite-sized Business Law is produced by We Edit Podcasts.
00:25:41.279 --> 00:25:45.839
Listen and subscribe to Bite-Sized Business Law on Spotify, Apple, or wherever you get your podcasts.
00:25:46.079 --> 00:25:56.000
The information in this podcast is for educational, informational, and entertainment purposes only and does not constitute legal advice, nor is it intended to promote the organizations with which our guests are affiliated.
00:25:56.160 --> 00:25:59.519
For more information, check out our show notes, and thanks for listening.