All right. Hello everybody. I know everybody's having great conversation around the table. Thanks everybody for being here at this conference. Scott for having us. I'm very happy to be here. Also the only news, and I don't mean like somebody negative on the stock market, I mean an actual bear in a decade. But what hasn't evolved here? Social media. A lot of lawyers. A lot of lawyers by the way, no cheering on that. Cheering, we have a lawyer. I also have a law degree but do not practice. So like girls saying I got my paper and I was free. So this is a great conference. A lot of heavy hitters here and I want to start with a disclaimer that none of you are going to be happy when this is over because all of you want something different from this gentleman right here. This is the Federal Trade Commission Chair Andrew Ferguson. Let's give him, whatever your political views, give a big round of applause. Because for those of us in public service, we are on United or Delta. There's a lot of other jets at the Aspen airport I saw which don't have a name on them and we're not, I'm not on those planes anyway. So, it's not the easiest place to get to, but look at this. It's absolutely stunning. And here's what I would call in CNBC an RBI. Random, but interesting. Chair Ferguson and I went to high school about 45, 50 minutes from each other on the I-81 corridor in the Shenandoah Valley, Virginia, Winchester, Virginia, Harrisonburg, Virginia. These are places that all you DC denizens have heard about but never actually been unless your kid goes to JMU. So that's what I just said. That's exactly, or you're on your way to UVA or the better school, Virginia Tech. So, I'm who wronged? It was a joke. Anyway, so we got a lot to get to. Let's jump right in. Chair Ferguson, thank you. We do have a little bit of breaking news actually in the media, which I want to get your comment on to start this off. A few minutes ago, Paramount Skydance filed a lawsuit against state attorneys general regarding the Warner Brothers potential planned merger. They're asking them to put up a bond of $1.8 billion dollars. Basically, under the Clayton Act, it's complicated. Let the lawyers sort it out. But there is, Chair Ferguson, some concern that if these state-by-state reviews of deals, Colorado, by the way, where we are, has one, if they are allowed to go forward, it could halt or at least hurt M&A, mergers and acquisitions. I know the news is fairly new. Your take?
Sure. I'm also pleased to be here and thank you for doing this. So I was a state enforcer and brought a lawsuit against Google when I was a solicitor general of Virginia alongside the federal government. So my perspective is both a federal enforcer and a state enforcer. And my experience in both of those says to me state enforcement is not a bad thing. And the best evidence of this is that when Congress passed the Clayton Act in 1914, they expressly gave states the power to enforce its prescriptions and to challenge deals. Now, that's been true for more than a century. And we have not seen an incredible, you know, even recently, we've not seen an incredible push among the states to challenge deals. I honestly don't think that's going to change that much for two reasons. One, states have pretty severe jurisdictional and resource constraints. They have to, you know, there has to be some sort of jurisdictional hook to their state in order to review any of these deals. For some deals, they'll have sort of like crosscutting national effects and states can probably get at those, but for most deals, that actually isn't going to be true. And so I don't think that this is, you know, proposed sea change. And number two, they have real resource constraints. I have them. Congress cut my budget by about 10% year-over-year. The president has asked for a substantial increase for the FTC. But the states have even more severe constraints and state attorneys general are appropriately focused primarily on stuff that is happening uniquely in their state. Frauds and scams that happen in their state, crimes that happen in their state.
But they're going after this merger.
Yeah. So those lawsuits against the deal are not just centered on those issues. They are going after what would be a national and international merger.
So there have always been some deals and some conduct that are sort of politically potent enough that the states are going to involve themselves. Virginia participating in Google against the United States in the adtech monopoly case is a perfect example of that. We devoted a lot of resources to that even though that was a national issue but part of the reason is because of the way people felt about Google. There are always going to be some mergers and some cases where the states are going to involve themselves and Congress wrote the law that way. I just don't think that for the mine run of mergers and for the entire merger ecosystem that state enforcement is going to be a substantial sort of like obstacle in the road. I could be proven wrong about this, but I've been in the states before. They have way more serious resource constraints than we're thinking and they also have higher priorities about the things that actually affect their states in a unique way and they have to answer to their voters.
So putting a finer point on it, same question asked differently, sounds like you are not concerned that these state AG's lawsuits against this potential merger will stifle or harm future M&A.
Future M&A. No, I don't want to comment on the specific deal because it's ongoing and the Department of Justice of course reviewed that deal. I just don't think that the, you know, the increased political focus on the left on state enforcement is likely to impose meaningful constraints on the M&A ecosystem. I understand why state AGs will want to say things like that. It's a potentially powerful thing to say on the stump, but the reality of resource constraints and jurisdictional constraints remain what they are. We've seen incredible deals. You could argue if we were sitting around and saying, 'What are the best five deals of all time?' I think you'd have to put Google buying YouTube for a billion up there, Meta, Facebook, then buying Instagram for a billion up on these lists. Would those deals get done? Would they be allowed today? It's always dangerous to answer hypotheticals. So, I guess look,
It's dangerous to ask them, too.
Yeah. Well, no, you get a free shot. Meta Instagram is maybe the easiest for me to answer because the FTC challenged that acquisition just, you know, several years ago. The first Trump administration filed a suit in 2020 about it. We continued to took the suit to trial and we're appealing it now. Again, you know, would that merger have gone through today? I would like to think that we probably would have had the same reaction to it that we're having today. But it is important to remember that merger enforcement is predictive. All of it is predictive. We have two companies or more depending on the situation trying to merge and we the enforcers have to make a prediction about the likely economic effects of that merger. The companies often have a diametrically opposed view. They have plenty of economists who are going to give the diametrically opposed view. We have economists giving our view. But at the end of the day, if it's a prediction, Meta Instagram, we remain quite convinced that the evidence after the merger bore out that that merger should not have gone through. I'd like to think that we would have made the correct prediction in 2013, but you know, it's always hard to know.
Do you see any mergers of that scale and size left to do when you look out?
Maybe in like cable TV. Just throwing that one out there.
Yeah. Fortunately, FTC doesn't do those.
But you have, you have a lot of say and weight in how things are viewed.
Yeah look I mean size is a relevant factor in merger review and I know that that's what makes it splashy on the front page of the Wall Street Journal but that isn't the primary focus of antitrust analysis in this country and frankly Europe says it's not. I think it might be but it's not supposed to be in Europe either. There are all sorts of other considerations that are supposed to matter. We won a merger trial, a billion-dollar merger trial on Friday. That was not a like mega blockbuster merger, but the evidence that we had showed that it was probably going to increase prices a lot for construction adhesives. And that, in my view, matters just as much as the mega merger because those are the ones that could otherwise go under the radar. They're going to increase input costs for constructing homes in America or in my case, fixing up a home like I do at home. And we won this merger trial on Friday, not to a ton of fanfare, but this is a huge win for the American people. And I think that it is as important or more important than the mega mergers that splash all over the front pages of the Wall Street Journals. And it wasn't about deal size. It was about market concentration and likely affects risk of coordination and things like that.
Well, you did just put out a comment on John Deere. I think you spoke to the corn growers association recently. Good Shenandoah Valley guy here. So we understand that that focus is there. There are some in this, I would say room but beautiful outdoor patio, that would argue that your sort of purview is not as objective as it could be. It's a little more subjective. How do you look at deals versus the objective x plus y equals z versus some of the things that got you into this position which is what some would view as subjective. You don't like it on this factor or that factor. They're almost extrajudicial. How would you respond to that criticism?
So, they're extrajudicial only in the sense that I have a $383 million budget and a lot of stuff I have to do. Remember, the FTC is basically two agencies in one. We are an antitrust enforcer and a consumer protection enforcer and I have to do all of that work with a $383 million budget. And so, generally what that means is, and there's another FTC chairman, former FTC chairman in the room right now, and I think you'll agree with me, you end up having to make very hard calls about where to deploy resources. Because if you had an infinite budget, you could look at everything equally. And so one of my jobs as a political appointee is to set priorities for how the agency is going to expend its resources. And I said upfront, I care primarily about health care, other inputs that affect Americans in their daily lives as consumers and participants in markets. And as participants in markets, that also includes their ability to speak online. And I think the way the FTC has spent its resources over the last 18 months is consistent with that. We've brought cases involving the advertising boycott and sued and settled with all of the major advertising firms in America about a boycott that they ran from 2020 to 2024 that was designed to prevent advertising dollars from going to disfavored publishers. We have brought all of our merger cases have been in health care or in home construction. And then we have a major conduct case that we brought last year that's also about home values. It's the Zillow Redfin consolidation that has to do with internet listing services. And so that has been our focus is lower cost for the American people on the markets they participate most. And for our economic markets that affect our political and social life like markets where people, where part of the participation is the ability to speak, make sure that those are not subject to illegal antitrust effects that affect both people's bottom lines and their ability to speak their minds.
Well, we have a lot of people in here whose platforms are the ability to speak. Meta, Google, TikTok, all the social media, they're all represented here. What is the right policy priority for managing speech and expression from your perspective online? These are trillion dollar questions.
It's not the FTC's job to police individual speech decisions made by Google, Meta, whoever. I see it. The FTC is having two roles, which is how we have expended our investigative resources. Number one is they can't coordinate on this stuff. Companies want to make independent decisions about speech on their platform. It is what it is. That's how the First Amendment works. That they can't get in a room together and coordinate about who gets to say what online. I think that the evidence in the Murthy against Missouri case demonstrated that a lot of that was going on. The government was facilitating a fair part of that for which I think every government official that was involved in that coordination should hang their head in eternal shame isn't happening anymore. But they can't get in a room and set the rules for the road that everyone's going to follow. We want competition on speech policies. And two, whatever your policies are, you have to follow them. You can't present one form of policy to consumers when they're deciding whether to use their platform and then have a completely different one in practice. That is exactly,
Don't want to talk about any ongoing investigations, but that is the core prohibition of our consumer protection laws. If you make promises to consumers, you have to live up to them. And that is no different for a company that is selling widgets than it is for a company who sells access to speech online. If you guys, if a company wants to say we allow this type of speech and you have these types of appeal rights and we will do this type of demonetization or this type of shadow banning, okay, but if you say we don't do those type of things and you in fact do, that is exactly what the FTC Act prohibits.
To get your eyebrows up and your lawyers engaged.
Don't want to comment on ongoing investigations. But I said from the drop, from the first week that I was appointed chairman by the president that we are going to examine to make sure that companies that make representations about their speech policies online are living up to them and that companies aren't coordinating with each other to set the rules for the road across all the platforms. If we're going to have a competitive economy, that requires that there aren't backroom bargains that deprive consumers of the benefit of competition. That is true with prices. It is true with the policies that govern speech online.
By the way, great job. Thank you very much. Suck in. They've adjusted the umbrella because we were getting, I was getting cooked, my fair Irish skin. So, thank you, sir. I really appreciate that. Also, stupidly wearing a wool suit. So, there's a one in two chance I'm going to pass out during this interview. So this is fascinating because it goes to so much that's happening in this room. I don't know, have you heard about this AI thing?
I heard a bit about it. It's a thing. It's in the news from time to time.
It is in the news. It's also at the TPI, Technology Policy Institute, because all the panels before us, AI in the economy, owning AI, quantum theory, etc. It's kind of a big deal. CNBC today had a story. It's a big one. If anybody here didn't see it, go check it out. Where like Walmart and others are like partnering up with OpenAI because shopping is now coming to AI, consumer protection, antitrust. There seems to be some sort of middle ground there between shopping AI and the FTC. What's the right way to look at a world that is faster than anything I've seen in 30 years evolving and dare I say taking over?
Yeah. So I think the general view I have is that Europe's approach is the wrong one. Comprehensive AI regulation before AI has really even germinated, which is what the AI Act in Europe was, is the wrong approach. A little regulatory humility is generally not a bad thing. The way that I see it from my little perch at the FTC is that I have a couple concerns. Number one is I want to make sure that the markets for the inputs for AI remain competitive. I don't want there to develop overnight sudden bottlenecked monopolies in the inputs for AI that deprive downstream users of the benefits of competition because someone upstream in the supply chain gets to jack up everyone's prices because it's enjoying a monopoly and maintaining it illegally.
Are you, sorry, are you talking about, because I know somebody talks about energy a little bit. Are you talking about the energy input costs, the construction input costs or what's spit out of AI in the sort of intrinsic cost from the technology itself?
Talking mostly about construction, energy, chips. I don't want either existing monopolies to use their monopoly power to maintain their monopolies in illegal ways. I don't want companies to buy up other companies that create monopolies somewhere in the supply chain. Look, from where I'm sitting, the LLM competition is like pretty ferocious. It's like almost a daily iteration of new adjusted LLMs, which to me is a pretty good indication at least there it's competitive. What I don't want is for prices to be rising because everyone that's trying to make an LLM has to pay a lot of money upstream in the supply chain. That's my principal antitrust concern. My consumer protection concern is that we have all sorts of companies, LLMs, application layer, all making all sorts of promises to consumers to get them to buy their stuff. I want those promises to be real. If you make a promise about something that your product is, that it can do, that it can help you with, that promise has to be honest. It has to do the thing that you're doing. And our consumer protection enforcement in the AI space has been focused on applications that say they can do something and they can't do that thing. The previous administration was taking the view that if an application has a potentially illegal use, like a use that could violate the laws that the FTC enforces, they're going to try to shut down that application. We undid one of those orders, made it very clear, look, the fact that an application can be put to a potentially illegal use does not make the application illegal. That's not how we're going to approach it. But if you have marketing materials that say our product can do this and your product can't do that. It doesn't matter what you're selling, AI widgets or not, it doesn't matter. The promise of our consumer protection laws is that when you make a promise to a consumer, the promise is honest. And we have brought cases against applications that have not lived up to their promises to consumers.
I'm going to not, not being political, I'll defend the previous administration in a sense that that was two years ago, three years ago. That might as well been 20 years in AI years, right? Because what I'm talking about is every week they're rolling out new models. Almost every week we talk about them on CNBC. You got Anthropic, you've got OpenAI, obviously you got Gemini, right? Meta's kind of doing their own thing. Do you feel like the playground, let's say it's sort of those four or five, if you threw in an Amazonish product, would those five be enough for fair competition?
I don't want to predict the future and I definitely don't want to say that a certain number of competitors is enough for competition and then in two years I look back and go I really shouldn't have said that. So I don't know. I do know that it seems to get easier and easier to build competitive LLMs. The cost that a lot of the big guys had to put in at the beginning in order to build their models. It doesn't seem like everyone has to put in the same costs to make new models or to challenge these models. That I think is a good thing. We also had at the very beginning of this AI revolution, and I do think it is a revolution, there was, you know, limited access to compute and so a lot of the AI companies had to strike deals with the companies that had hyperscaling in order to even grow their LLMs but now compute is also growing. I mean look, I'm just going to give an example of this. During inauguration and the weeks leading up I was taking meetings all over the place with all sorts of big business CEOs, small business founders to try to figure out like what is happening in the AI ecosystem. Should anything I should be worried about? And what I heard early on was we have this hyperscaling switching problem where I buy access, I buy compute from a company and then I have new compute needs and I want to move my data from one hyperscaler to another and they say sure but we're going to charge you an exorbitant exit fee. And I went all right well that does sound like a problem. And I mean within like three months we had people like talking to the same guys like no that's not a problem anymore. There are now all sorts of tools we can use where I can just throw it on all the hyperscalers, access it whenever I need it. This is actually not that big a deal anymore. And I sat there and thought gosh if I had spun up like an entire investigation on the basis of this thing by the time I got around to doing anything about it would have passed me by. It would have been an eon in the development of these products. It's,
Did I just date myself on that? Yeah. He's nodding. I don't look, you were about the same age. I think now you're much younger.
But you remember it. And people thought by the time it was decided, it was moot. Yeah. Right. GM US Rails, if you remember that back in the 30s, General Motors was sued for like buying up all these trains and then, you know, basically taking the tracks out. They were going to modernize them, but the company was like a secret holding company for GM. They just wanted to sell you cars. I think the court fined him, remember, I didn't go to UVA law school, so forgive me on this one. I think GM was found guilty, but they fined him a dollar because the judge is like, 'Well, it's too late now anyway. Nobody cares.' Is that, I mean, that's how fast AI is moving. So, how does the FTC, how do you try to figure out where the puck is going or do you just kind of react to where we are now because that's in your view your judicial purview?
Yeah. So for like the LLMs themselves, I'm not going to try to predict where the puck is going and instead I'm going to focus on what the FTC knows from experience, which is the industrial part of the supply chain that makes the LLMs possible. That is,
The concrete, the electricity,
The chips. I mean that is the type of thing that antitrust enforcers are accustomed to, the supply chain, as opposed to trying to predict where a hyper, in my view pretty, at least as of today, competitive but extremely innovative and extremely fast-changing market at the LLM and application layer is going. Focus on sort of like the meat and potatoes part of the supply chain that makes those applications and those models possible, that we can do. That's just ordinary industrial organization economics and antitrust. And yes, it's also going to be fast moving because the supply chain will be wanting to keep up with the models and the applications, but that's the type of thing that we can look at and be like, we've seen this type of thing before. We know when consolidation there becomes dangerous. We know when agreements are likely to raise prices or reduce competition or cut off innovation as opposed to trying to get out in front of the AI developers don't even know where this is going. It would be insane for a regulator to say, I know where it's going and I'm going to make predictive regulatory choices on that basis. But further back in the supply chain that we can apply ordinary antitrust to.
Like I'll just make a wild hypothetical like if Nvidia tried to merge with I don't know like ARM or Kroger and Albertsons.
I just make randomly pulling out companies out of the hat. Those are real examples by the way.
I think if Nvidia and ARM were to merge we would have to take a very careful look at that.
That is antitrust enforcer speak for you know we would have concerns. Yeah. Because of the ultimate input cost to AI which would then trickle down to what consumers ultimately pay.
Yeah. And not just AI. I mean look, yes, we all talk about Nvidia right now because their GPUs are necessary for the models to do their thing. But there are a lot of consumer applications for chips that don't have anything to do with AI. Look, I build my own desktops at home mostly for fun. And like a year ago, I decided that my rig needed to be updated. And I was like, 'Oh, well, you know, RAM is pretty cheap right now. Like, maybe I'll do that.' And then the job got busy and I was like, 'I'll take a look at this later.' In six months, I looked at RAM prices and went, 'Oh my god, I'm not building this right now. This is crazy.' And there are consumer applications for all these things that matter independently of AI. So yes, does AI matter? Yes, it matters in large part, I think, not just because innovation is cool. Innovation is cool, but we can't lose this race with China. And I think that our main competitive edge over the Chinese is that our economy is competitive. And that yes, competition breeds chaos and it can feel uneasy at times, but that is far likely to generate a winning formula in this race than a command control economy. But we have to make sure that we keep all of the inputs to that competitive or we're going to have a slowdown at the very end.
of the supply chain, too. But not just companies. I don't want consumers to have to pay way more for chips than they have to for all sorts of other applications, including the phones that we have cast all about us.
Yeah. Whose prices also keep going up, by the way. I'm not sure what goes up faster, iPhone or phone prices, not just iPhone, or coffee in Aspen. Those two things every time I come out here it's like an extra dollar. Um, but you know, you go to that. This may be a little off topic, but like you look at the car industry, right? I was just in the upper peninsula of Michigan last week. I've got a lot of friends from Detroit. They're terrified of Chinese competition in part because they can't compete on the labor cost. Period. End of story. AI the same way, right? The Kimmy model, Deepseek, these things. How do we regulate so much of what we're doing now? Now we're trying to figure out how to manage through AI tokenization costs. It's all there. And then you got this Chinese models over here where their input costs the stuff you just talked about. There's 0% chance that we can compete with those just like automobiles because we need to pay our people more than $3 an hour to make a car, right? How do we manage that from a consumer protection, antitrust perspective, your purview? How do we do that with AI? Because there's no freaking way we can compete on cost. Zero.
