Amazon drops data center NDAs, and AI agents want your credit card

Amazon says it will stop using NDAs when negotiating data center deals with local governments, following a similar move from Microsoft earlier this year. Secrecy has fueled community backlash against AI infrastructure, with opposition leading to hundreds of proposed and enacted moratoriums from New York to San Francisco. Meanwhile, a wave of startups is betting that consumers will hand AI agents access to their inboxes, files, and credit cards — if they can get websites to let them in. On this episode of TechCrunch’s Equity podcast, Anthony Ha, Sean O’Kane, and Rebecca Bellan dig into whether ditching NDAs will change how data center deals get done, why trust may be the real product for personal AI startups, plus a couple of deals of the week. Listen to the full episode to hear more about: The White House’s plans to make plans with its new Super Intelligence Force Ghost’s $3,499 computer built to run AI agents around the clock, and why the Equity crew doesn’t see it going mass market Tab, Underdog, and Hark pitching privacy as their product, plus the business model that could turn your agent into a salesperson United, Apple, and Amazon putting up walls against outside agents Lambda’s reported $4 billion raise, and the single customer behind $35 billion of its backlog Uber’s $2.3 billion all-cash deal to get into the office catering space with ezCater We’ll see you next week at TechCrunch Disrupt 2026! Get 25% off your tickets with code Equity25, and we’ll see you on the Builders Stage Tuesday at 9 am. Subscribe to Equity on YouTube, Apple Podcasts, Overcast, Spotify and all the casts. You can also follow Equity on X and Threads, at @EquityPod. Chapters: 00:00 Intro 0:55 Trump’s “superintelligence force” and the great AI rebranding 3:19 Amazon drops NDAs on data center deals 9:17 Why the AI industry has a communication problem 12:51 Personal AI agents: finding money in your inbox 15:27 Ghost’s $3,500 computer built for personal AI 19:16 Trust is the product: privacy-first agent startups 27:36 Lambda’s $4B raise and the neo cloud backlog problem 31:22 Uber’s $2.3B move into catering 35:10 Outro Learn more about your ad choices. Visit megaphone.fm/adchoices

Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.

• 0:00Speaker 00

  This episode of Equity is brought to you by Schneider Electric. Schneider Electric, your energy technology partner.

• 0:05Anthony Ha?

  Amazon says it's done with NDAs for data center deals as community backlash continues to grow. Also, startups are getting into the personal AI agent race, as one startup says it's selling $3,500 computers built for personal AI. And of course, we're keeping our eyes on a couple of deals of the week, including Uber's $2.3 billion move into catering. Stick around.

• 0:35Anthony Ha?

  Hello and welcome back to Equity, TechCrunch's podcast about the business of startups. Today is Friday, October 9th. I'm Anthony Ha, and I'm joined by TechCrunch's Senior Reporter, Special Projects, Sean O'Kane, and TechCrunch Senior Reporter, Rebecca Bellan. And Rebecca, you have been on vacation, so first of all, welcome back, and I'm sorry. Did you see the news that Trump announced a new super intelligence force?

• 1:01Rebecca Bellan?

  Yeah, I mean, of course, that happens while I'm away. But from what I, from my catch up, I can tell that it feels like it's kind of a nothing burger right now. It's just like plans to make more plans in the future, maybe? I don't know, do I have that right?

• 1:14Anthony Ha?

  Yes, that's.

• 1:16Speaker 00

  It feels like an AI agent could solve this.

• 1:19Rebecca Bellan?

  Yeah, I mean, well, it's, uh, it's interesting. So now we're calling it super intelligence, right? And then, did you guys see at the time of this recording, Trump posted on Truth Social that if you're calling it artificial intelligence, not super intelligence, you're, you're an enemy, so.

• 1:35Anthony Ha?

  Although I think Elon said that SpaceX, you know, the, the XAI division within SpaceX is in fact going to become XSI, or something like that.

• 1:45Rebecca Bellan?

  Yeah, he's changed it to SpaceX SIs.

• 1:47Speaker 00

  SpaceX SSI, yeah, rolls right off the tongue.

• 1:51Anthony Ha?

  Yeah, I mean, we, we talked about this a bit last week, not so much the, the force, but the, the term. And I think part of the, the bigger story here, besides, you know, the super intelligence force, which is unlikely to do a whole lot, is just the fact that the kind of outcry, the, the backlash towards AI, which I think comes from a lot of different corners, has gotten loud enough that the Trump administration feels like it has to kind of do something. And, and so it's kind of doing this kind of weird rebranding, pretending, kind of nodding towards safety exercise.

• 2:28Speaker 00

  Listen, we're, this is, a lot of this is artifice. The, you know, super intelligence force is like more shiny titles for people to add to their, their LinkedIn or whatever. It's like, okay, cool.

• 2:41Rebecca Bellan?

  Yeah, it feels like it's like part and parcel of this great rebranding we're seeing about AI, because everyone's starting to freak out about it. So they're like, no, no, don't get too scared. Like, it's, it's not artificial, it's, it's super intelligence. And here, look at these cute little AI personal agents that are gonna help you get money back when they look through your email, which we'll talk about later. But it's interesting. And then I guess also, this week, at the same time, I don't know if this is how this, how related this is to the super intelligence force, but there was a, a science summit, right? Where they announced $6 billion in initiatives for using AI in science, I guess. That's, you know, 2.4 billion in tools and compute credits from companies like Nvidia, AMD, OpenAI, Anthropic, and Google. What do you guys make of that?

• 3:25Anthony Ha?

  Yeah, I was just reading the, the announcement before we started recording, and it felt like, first of all, just sort of the very puffed up language of the announcement, like, this is the greatest investment in science in America's history, which, uh, you know, I think it's fair to say that the Trump administration has not been the greatest friend to science in American history. But I'm sure we're gonna continue talking about a lot of these bigger issues with AI. In fact, I mean, this is also gonna thread through the rest of the episode, spoiler. But there's probably gonna be more developments, and just in the next few days. And so just as a last, final, final reminder, Disrupt is next week. As we record this, we're gonna have a live recording of Equity on Tuesday at 9:00 AM.

• 4:09Anthony Ha?

  And if you use the code EQUITY25, you will get a 25% off a ticket, so that's EQUITY25. Let's keep talking about AI and the AI backlash, and specifically now, data centers, which I think have been kind of one of the main focuses of, of, you know, community resistance to AI. And Amazon put out this blog post at the end of last week, basically, I mean, it was really trying to sort of make the case that data centers are great. And then kind of buried in there was also this announcement that they're no longer going to be using NDAs in the process of their negotiations with local governments. Sean, what did you make of this?

• 4:51Speaker 00

  I thought it was interesting, I, first off, good catching this. It's a interesting thing that it was sort of buried in the blog post, and part of a larger story that Amazon's clearly trying to tell. But I think it's really important. I don't, I mean, I guess I understand why they weren't putting it in the foreground, but to me, this is at the root of so many of the problems that the tech industry currently faces when it comes to AI broadly, but data centers in particular. You know, there are so many things that could be said about whether or not people should be worried about, you know, the effect that data centers have on their property value, on their electricity bills, on their water usage in the area, and whether or not those

• 5:34Speaker 00

  things are real problems, inflated problems, like, et cetera, et cetera. I think so much of that just really goes back to the fact that a lot of these projects are being negotiated essentially in secret. Sometimes they're being negotiated with it known that it's a data center, but just not knowing who's actually, you know, going to use it or occupy it at any point. And to me, like, this is, I mean, this is a longstanding issue with the tech industry, right? Like, I always go back to the mid 2010s, when Uber was expanding really fast. Uber was one of the companies that really used this tactic a ton when it came to working with local governments in particular. I remember there was a, I was working at The Verge at the time, and we had a freelancer

• 6:19Speaker 00

  write a story about a city outside of Orlando, where Uber had, you know, was trying to negotiate basically tax subsidies to subsidize rides for people to use public transit after the Uber trip. And the whole thing was negotiated in secret. And not only that, which was frustrating on its face, but also, like, there were all these, because of the secrecy, you know, you could just see Uber getting really comfortable with asking for a lot of other stuff, too, because they figured it would never be known, like, literally asking the city not to put Uber's logo on, like, trash, or even on food, because like, it could be sort of related to being thrown away, or like, winding up in a toilet. You know, it's just insane stuff. So like, there needs to be more of this.

• 7:03Speaker 00

  I'm glad Amazon's doing this. We'll see if it actually does have any positive effect, but, you know, people are pretty entrenched right now.

• 7:10Rebecca Bellan?

  I'm curious about what effect this will have, and whether it will lead to communities successfully delaying or, or killing any infrastructure that the companies need to build, right? Like, I think, from what I've heard from some people I've spoken to who deal with data centers kind of every day, a lot of these moratoriums that are getting, you know, passed, and, and there are a significant number of them, including, I think there's one recently in, in San Francisco, right? Which is pretty ironic, but they're only for like, a year or two years. But these data centers aren't coming online anyway for the next two, three, four, five years. Like, all the, all of the build outs are, are going to be massively delayed for a number of reasons, from zoning to funding.

• 7:51Rebecca Bellan?

  You know, of course, if, if the pushback keeps going, then it might have an effect, but I often wonder just how much of an effect this political pushback is having on this.

• 8:01Anthony Ha?

  I mean, you know, yeah, it's certainly worth noting that, I think the, the moratorium I'm most familiar with is the one in New York, where you and I live, Rebecca, and, and I believe that's a, a one-year moratorium, and that's on permits for large projects. So obviously, that's already, there's, you can see that there's some wiggle room within that, and, and often they're sort of framed as, hey, we need to take, it's not like, we think this is bad, let's not do it. It's, we need to take the time to study this, which gives them a lot of room to say later on, actually, we've decided this is great, we love it.

• 8:33Anthony Ha?

  But you can tell that the people trying to build these data centers are worried, which is why you have, you know, these kind of unhinged tweets from people like Trump or David Sacks. Only time will tell in terms of like, what, how long and how deep that resistance is. But, but I do think they are worried. It's, it's also noteworthy that I think Microsoft, earlier this year, made a similar announcement about not using NDAs. You know, there's, there are all these other criticisms, right? I don't think that getting rid of the NDAs automatically is gonna make the people who oppose these data centers be like, great, let's, let's do it, I'm in. But it, I think a lot of that, kind of just the beginning of that conversation is really like, well, we can't, we don't even know what the heck you're doing, because you've, you've basically, like, negotiated all this stuff in secret.

• 9:18Anthony Ha?

  So it does seem like a good first step to winning people over.

• 9:23Speaker 00

  I mean, this is the larger problem here is just a communication problem, right? This is a problem, I think, that goes across the entire tech industry. But it's especially acute now because it's the, you know, we're living in the era of going direct and you don't need to talk to the media, the media is dead, et cetera. And I'm not like trying to beat that that horse because like we're in the media and think we're still relevant. But I think I think we've seen through a lot of the controversy and backlash to AI and data centers that the industry just does a really poor job explaining itself and that, you know, they are obviously and constantly frustrated with the fact that most people can't envision a world in which this stuff is useful for them.

• 10:07Speaker 00

  And that's in part because a lot of times the ability is not there and in part because they're just not explaining it well. Like I wrote a story last week for us about how it's been a year since we've really heard Tesla beating its own drum about building towards what it originally was, quote, "sustainable abundance" and is now "amazing abundance" as its mission statement. And it's just wild to me that we're a year into that and the company still doesn't have a very good definition of it, especially considering that like they released this whole sort of, you know, multi-page PDF laying it out when they announced this and even Tesla's biggest fans criticized Elon Musk and Tesla for how unspecific it was because it was,

• 10:50Speaker 00

  I mean, it read like LLM nonsense and they just, they still don't care to go through and explain this to people about like where this, and you know, and I, I think that's crazy to me because we are starting to see capabilities here like OpenAI, you know, kind of like one-shotting like 500 math problems this week in different ways. Like there are things that these models are capable of now that you could probably try to draw a line to people. But, you know, I think the, the fact that the abilities weren't there, that the communication wasn't there and that initially a lot of this stuff was really shoved down our throats, co-pilot, you know, I, it's just there's, they've got to really kind of talk themselves out of it.

• 11:34Anthony Ha?

  Yeah, I mean, I, one of the things that this whole thing has kind of made me think about is just the narrative and this sense that for a while, it really felt like the, it was just like you said, kind of being pushed down our throats as this idea of this is inevitable, this being sort of, you know, this kind of AI-driven future and get on board or you're just going to be left behind. And I think because, you know, obviously that future is still kind of barreling ahead in a lot of ways, but in a few key ways, there have been these obstacles. I, what I've really liked about that is that this sense of no, like you actually have to make the case for it. You can't just say, this is the future, shut up. Although I think that's actually what, you know, again, the Trump administration and other folks are trying to do is just be like China, China, China, China, China, don't argue with me.

• 12:16Speaker 00

  Yeah.

• 12:17Anthony Ha?

  And then specifically about the, yeah, the Tesla and abundance thing, Sean, I, yeah, I really liked your story, although what was shocking to me is not so much that they haven't come up with a definition of abundance, but that that was a promise that they were making in the first place. Like that is wild and I think speaks to how strange our relationship to Tesla and particularly the people who are invested and kind of like ride or die for Elon, that it's not like about like, is this company going to sell a lot of cars? It's, it's really about sort of like this vision for remaking society, or at least that's kind of what they're being sold.

• 12:52Rebecca Bellan?

  Yeah, it's always been the vision.

• 12:54Speaker 00

  Yeah, it's like, we're going to kill money.

• 12:55Rebecca Bellan?

  Yeah, it's always been that. Whenever I talk to people who back Elon no matter what, and I'm like, but the numbers. And they're like, no, no, you don't understand. It's his vision for the future. And I'm like, right, yeah, cool, cool, cool, cool, cool. The future is now, right? I mean, there's all these companies now promising for, you know, to give us, they've been talking about personal assistance and how AI is going to help you with your everyday tasks for a while. And now some of that's starting to come to fruition and you're, you're, rather than the broad pushback people have been giving to LLMs for, you know, AI slop writing and stuff, people are like, oh, wait, it can find me money in my inbox? It can book my flights for me? Oh, wait, okay, let me, let me see about that. So I don't know how good any of it is yet. I'm testing out a few.

• 13:39Rebecca Bellan?

  I don't know about what you guys think about it, though.

• 13:42Anthony Ha?

  What moves our world forward? Energy. History has been shaped by it, revolutions won with it, and progress dependent on it. For 190 years, Schneider Electric has been a leading force in this story, creating the technologies that harness energy and transform the competitive landscape. And today, in the dawning era of energy and industrial intelligence, Schneider is helping customers build more efficient, resilient, and sustainable operations across every sector, from data centers to homes, from buildings to industry. Schneider Electric, your energy technology partner. Discover more at se.com.

• 14:18Anthony Ha?

  Yeah, I mean, last week, actually, Sean was, I can't remember if it was last week or the week before Sean was telling us about, that Muse did actually find some, some money for him, but he wasn't as grateful as, as I would have thought. He was like, that's cool, but like, what have you done for me lately?

• 14:33Rebecca Bellan?

  Wait, but how much money? I didn't hear it.

• 14:35Anthony Ha?

  Let's be, let's really get into it.

• 14:38Speaker 00

  It found me, it found me a good chunk of change in some unclaimed property in Texas that, you know, you know, it was one of those things that's, you know, I should be, I should have thought to look at that myself and it did that for me and that was fantastic. And, but that's like a one-time thing that, you know, I don't have the ability to do that again, right? Like, it's just, it's sort of, and, and, you know, I think now we're talking about agents, like, this is the first real sort of stake in the ground that I think we've seen this industry place as far as trying to correct what I was just talking about and communicate to people what might be useful about this technology that goes beyond

• 15:21Speaker 00

  it being better Google or like psychosis inducer. And I still think they're struggling with it, right? Like, I don't know about you guys, like, I understand why so many people in Silicon Valley think that agent stuff is like, is it for people who are not using it at an enterprise level? But for most people, I feel like, you know, from my view, they'll have this, a similar experience to me of like, oh, I can tick off a couple of these things that it's suggesting to me right out the gate, that's helpful, but like, will there be enough of a business around the more passive uses of something like Muse where, you know, it's not commercial

• 16:05Speaker 00

  and it's really more about like finding information for you or something like that? I don't know.

• 16:10Anthony Ha?

  Well, I also think that could be a good segue into our second big theme, which is, of course, is AI-related because it's 2026. But it's basically that there's also, you know, a number of different startups trying to take advantage of this as well, kind of figure out where the opportunities are that haven't been claimed by the OpenAIs and the anthropics. And, and we can, we can kind of talk about whichever one of these like recent companies interest you, but maybe we could start with Ghost, which is a startup that, you know, got written up on TechCrunch and, you know, from what I saw at least, the story really seemed to blow up for us. It really seemed to strike a chord. And this is a startup that is building a 3,499 computer for personal AI.

• 16:52Rebecca Bellan?

  Yeah, so it's, it's $3,500 for a computer. This is core and the entire point is to run its AI agenda for you. So it's got an Nvidia GPU in there, its AI model's already installed. I think they have their own software browser file system and it can run continuously without needing the cloud. So, yeah, I think, you know, how is this different from a MacBook Mini? I don't know. I was like, is this just an external hard drive with a little bit of extra? Do you, you still need a phone plus the app to like make this thing work? And then I imagine one day they'll, they'll be like, well, you can't have two devices, let's just put it all on one. And then you've just what recreated like the phone and the computer? I just, yeah, it's interesting.

• 17:31Speaker 00

  I mean, it's a, I think people like it because it's hardware and people like hardware and, and dedicated hardware. We saw-

• 17:37Rebecca Bellan?

  Hardware is so hard.

• 17:39Speaker 00

  It is hard, but like people like the ideas that go into it. It's, you know, an interesting design. It looks like a, you know, kind of a PS5 meets like the monolith from 2001: A Space Odyssey. But again, we run into a similar problem here where- You know, one of the people that's working on this said that like, you know, this is an example given in our story, it's supposed to, you know, maybe it will proactively send you a notification that you might be getting sick because it sees your resting heart rate is elevated, sees from computer history that you are up late working on a project and has seasonal context that it is flu season in San Francisco. It's like, do we need a $3,500 box to tell us that? And it like, some people I think do, like, some people are looking for that thing that like lets them

• 18:24Speaker 00

  externalize what they're thinking sort of subconsciously. And I, not to say there are other uses of this that won't be useful, but you know, to me, it's sort of like, we're still, we're still so early in the like, app only phase of agents and having it on your phone. And we haven't really found a bunch of traction with people, you know, outside of this sort of Silicon Valley set. And this is clearly targeted at them. And I know that there are a bunch of people like this is running a bunch of open models like Quen. And like, I know that there are people like, have been scrambling to put together their own setups in their apartments or houses that resemble something like this without the polish. So I'm sure that they will

• 19:06Speaker 00

  find a market for this among the, you know, the sort of industry folks.

• 19:11Anthony Ha?

  Yeah, my take was that this is definitely not mass market product. Like, normal people are not going to spend $3,500 for an AI computer that, at least from, you know, it seemed like on our coverage, we were trying to ask them some questions just about like specs and things. And we're not getting a lot of detailed answers. It was just sort of like, look, AI computer, you know, I think there is an audience for that. I mean, and, you know, the most of them live in Silicon Valley and spend a lot of time on X and are very excited about this. I, you know, presumably, you know, you raise, I forget exactly, I think it was like $11 million. Not a huge round, but like, at that point, you have to, at some point, think that there's

• 19:53Anthony Ha?

  going to be a bigger market than, you know, hardcore AI evangelists in Silicon Valley.

• 19:58Rebecca Bellan?

  Well, you know, it's not, I think that one thing that we're seeing, because there's a few other startups that will, that we'll mention that are playing in this personal AI assistant space. And one, one like common thread throughout them all, for the most part, is what I see is, you know, being able to run something locally on your own device, and also having this like, privacy focused thing, right? Because it's like, these apps, these agents will not work unless they have like, all of your access, right? Like, if, you know, I don't know if either of you have ever been an assistant, I was an assistant, an executive assistant, I had documents of, you know, what my, you know, boss liked, what her constant, like to go orders were, like where she liked to sit on a flight, I knew, like, so, like,

• 20:42Rebecca Bellan?

  honestly, this was all in handover that I passed over to someone else. I was the agent, and I had her, you know, SkyMiles, her credit card, I had newer kids fricking social security network, right? Like, I had all of it. Right? And so I think that this is, you know, if you're going to give that to an agent, then you're going to need some kind of assurances of privacy. And it's, it's difficult because it's like, on the one hand, do you really want to give all of that to Meta? Right? Do you really want to do that to give it to Muse? Do you want to give all that data to Muse? On the other hand, do you want it to go to a company that has the infrastructure set up to protect your data versus a one-year-old, one-month-old startup that's using like a shared cloud? Right? So it's, it's interesting. I mean, there's some

• 21:26Rebecca Bellan?

  others like Tab, they just emerged from Stealth, that's $300 million valuation, they haven't disclosed their funding info, but they're also kind of focused on that. And one of the excellent, this was written by Dom, one of our reporters, excellent kicker quote, I thought, capability is table stakes, trust is the product, which I thought was pretty cool.

• 21:46Anthony Ha?

  Yeah, and it did seem like trust was a big theme in a lot of these startups now. There's Underdog, which is also supposed to be kind of emphasizing privacy, because it runs entirely locally, which I'm sort of curious, like, what that actually looks like, how powerful a computer you have, how powerful the, the model can actually be if it's just running entirely locally.

• 22:07Rebecca Bellan?

  Maybe they'll teem them up with Core.

• 22:09Anthony Ha?

  Yeah, exactly. And then there is Hark, which is also focused on privacy. So that does seem to be a recurring theme. You know, I would say that, you know, certainly personally, that that idea of either giving an agent access to my hard drive or my credit card, I mean, I'm, you know, and I'm probably a little further behind most of our listeners, but like, I'm not ready to do that yet. I'm not going to do that. And then part of the challenge then is that seems to be where they think a lot of the kind of value is, right? Oh, man, once you do that, like, all these amazing things are going to happen. Once in a while, we'll delete a critical file, or buy something really expensive that you don't want, but it's fine, don't worry about it. So, uh, I, yeah, I'm not very confident they're going to get me over the chasm at least.

• 22:54Speaker 00

  Yeah, I mean, I think Rebecca brings up the most salient point here, like, from where we stand with the rollout of this stuff, which is, everybody's kind of stuck between a rock and a hard place, right? To get more out of these agents in the way that these companies want you to think they're, you know, the, the top capabilities of what they're able to do, you have to give them so much. And you either have to trust a company that has that infrastructure, and to some extent regulatory oversight, but a bad history of letting, you know, that data get misused in some way, or a company that only just started existing a few weeks ago.

• 23:37Speaker 00

  And it's, it's just like, that's not a decision that I think most people want to make right now. The other thing is, like, I know with Underdog in particular, I was, their whole thing is not only it being like local and trust-based and, and keeping it, you know, they're also changing the business model around it. And I think we're seeing elements of this with Muse coming from Meta, but the idea being that like, they don't want to use the information about you to target ads about you or anything like that. It's really more about taking small cuts and working, you know, in particular with Stripe, which is, I think, you know, one of the colistens is an investor in Underdog, to take a cut of transactions that you're going to have this

• 24:19Speaker 00

  agent do on your behalf through Stripe's sort of processing, which is interesting and, you know, reassuring to some extent that like, they don't want to use or abuse all that information about you. But then again, if that's your business model, like, your agent that you're building is going to be sort of inherently incentivized to be pushing me to be buying stuff all the time, because you want to be making the money off of those transactions. And like, I don't know that that's what I want to use agents for. Like, we do not have as much money as coders in Silicon Valley, right? Like, it's just not the, like, again, there's this skew of like, what all the incentives are, sort of from step one. And so, I, good luck.

• 25:03Rebecca Bellan?

  And I'm also, I'm also like, okay, good luck trying to make a business model that doesn't include advertising at this stage. Like, this is what, like, everyone tries to not do advertising, and then they always have to do it, because, like, we're seeing this with OpenAI, because it's so expensive. Like, I'm hearing, I'm hearing from people that they're, these models, I mean, it's agentic, right? Like, they're, they're running like, hundreds of thousands of tokens for like, one simple task, or like, not even a task. And so it's just, it feels like it's going to get really expensive for these companies. I question whether they are a little more than just like, again, wrappers for agents that aren't their own. Like, I'm not sure, I don't think any of them are building their own models. I would assume they're not, they're probably just using like, anything off the shelf, anything that's open. And so, yeah, I think that things are going to get really expensive for them. And I don't know

• 25:46Rebecca Bellan?

  that just like a little take rate off of Stripe is going to be enough for, you know, getting, reaching any form of profitability. But then I also wanted to mention before we move on from this topic, I think it's interesting, the timing of this all. It's really ironic, right? It's like, just in like, we're still in the weeks, we're still all recovering from the OpenAI hugging phase, the rogue agents, like, the idea that, you know, agents can perform activities that could be catastrophic, that you really don't want them to perform. And then they're like, but, but here, like, have an agent take over your personal life? Like, what? Like, that's, it seemed like the riskiest time to do that when we don't know that they're aligned. These companies, they're selling these privacy solutions that they simply cannot guarantee,

• 26:29Rebecca Bellan?

  because they don't own the models, they don't know what the alignment is, right? And it just, it feels like, it feels smart to me. Like, the timing actually feels, in an evil genius way, kind of smart, because we're, once we get to a point of like, peak convenience, I think a lot of the people who are scared of AI and saying, regulate it, like, take it down, stop, are going to say, yeah, it's all right. Like, it's, well, it's made my life really easy. So I guess, that's fine.

• 26:55Anthony Ha?

  Yeah, I guess we'll see. I mean, I, one other thing to mention on this topic is also that, when we talk about trust, it's not just about winning over the trust of consumers, or even of enterprises, but, you know, that there are these partners that also need to, quote unquote, work with the agents in some way as well. And so we've had some stories in the last week about, you know, Apple, like, limiting how it provides disk access to, you know, AI agents, and then also a number of websites that are now also making it a little bit harder for, or I don't know if they're making it more difficult, or just those, people are discovering those obstacles more, that like, you know, people are trying to have Muse go and buy plane tickets for them, and United is like, we, no, we don't allow that. That's not something we support. And-

• 27:41Anthony Ha?

  So I think that probably there's going to be some level of change on that. But I think it's also fair for, you know, the Uniteds of the world to be like, hey, hey, with like, you don't get just sort of unfettered access to like, buy plane tickets on our website. Half of them are going to like, end up being like refunds because people like, did not want to do that. And so I think there is going to have to be this earning of trust, this negotiation period.

• 28:04Rebecca Bellan?

  I was just gonna say, Amazon did this. That's it. That's, Amazon did this with like perplexity and stuff. They were like, no, you guys cannot come, like, use your agents to come and shop on our platform. Everyone's got to use Rufus, or now it's Alexa for shopping. But anyway, we do need to move on, so.

• 28:18Anthony Ha?

  Right. But no, this is a great reminder, of course, that like, safety is important. Safety is often, sometimes, an excuse for other kinds of protective behavior. But let's talk about some other startups. First up, we have Lambda, which I think is one of the NeoCloud startups that we've talked about before. Rebecca, why don't you tell us what's going on with them?

• 28:39Rebecca Bellan?

  Yes, this is a NeoCloud AI computing startup. This is reportedly raising. So the Wall Street Journal had this scoop. This would be a $4 billion, opt to $4 billion round at a $14.5 billion pre-money valuation. And this could be their last private round before their 2027 planned IPO. I myself am skeptical if they will make it to that IPO because they were meant to IPO this year, and I think even at the end of 2025. So, you know, this could be one of those like, buying time situations. But, um, yeah, it's, uh, it's quite the raise.

• 29:15Speaker 00

  I think especially companies like this, but broadly speaking, you know, we talked about last week, Aura pulling its IPO. But I think we're waiting to see how the Anthropic one goes, to really know how many of these companies are going to go through with this. And sort of conversely, you know, it seems like OpenAI is about to close a $30 billion funding round. And if there's still that much liquidity in the private markets, then, you know, those two things could really incentivize people away from the IPO market next year.

• 29:49Anthony Ha?

  Yeah, I'm also curious about sort of what the appetite will be specifically for NeoCloud companies, because I think on the one hand, you know, they, they have like revenue and real revenue in a way that, you know, some of these other AI companies do not. But then, you know, what are the long-term prospects for that? That, you know, obviously, there's just like intense data center build out. I mean, with some resistance and some slowdown as well, so maybe a little TBD on that. But, but potentially, you know, they're selling something that's really valuable now, but might not be in a couple of years.

• 30:18Rebecca Bellan?

  Gosh, you hit the nail on the head. So I think in the short term, I think these companies are, there's a lot of frothy appetite for them because they have the capacity that everyone needs, right? But, you know, in the long term, it's hard because, so, the Journal had, you know, reviewed a letter to investors that showed the backlog grew from 15 billion in June to 50 billion in September. That looks like a big deal, but 35 billion of that appears to have just come from one customer, which is Anthropic, which signed a deal with Lambda in late August. So a lot of this is based off of like, a lot of the NeoClouds have the unhappy problem of having like, one or two major customers that they're basing everything on. It's not revenue, it's backlog.

• 31:00Rebecca Bellan?

  Like, that money's not coming in for a long time, and those are contracts that could be canceled, right? If the demand doesn't show up for whatever reason, right? So, yeah, it's, it seems like potentially risky there. And then, I just also want to float that Lambda is potentially raising this equity round, but that's on top of something like four separate debt raises in the last, you know, few months or something, or the last year. I, it's, it's, the data center builders are, build outs are largely funded by debt. And you're seeing this kind of trend, or I'm noticing this trend among NeoClouds as they prepare to go public, where they are raising an equity round that makes them look

• 31:43Rebecca Bellan?

  legit, and then they're raising a bunch of debt to help fund the build out. You're seeing this with NScale, which also recently raised, they raised, I think, like $2 billion earlier this year, and then they've put on another like, two or three rounds of debt, including one that TechCrunch reported on last week. So, yeah, there's a lot of moving parts here, a lot could go wrong.

• 32:04Anthony Ha?

  Yeah, we'll keep an eye on that. In the meantime, we do have one final deal that I'm excited to say that, as far as I know, does not have an AI angle, and that is Uber getting into catering. Sean, what's going on with that?

• 32:16Speaker 00

  I can make it AI if you want.

• 32:19Rebecca Bellan?

  No, please.

• 32:22Speaker 00

  I actually, yeah, I mean, no, I won't. Uber is acquiring a company called EasyCater for $2.3 billion in cold, hard cash. Good for, good for the catering business. I'm sure they use some AI because they are not like a caterer. They don't come and like, put the food on your, your, your sort of six-foot table. They are like a marketplace kind of matcher for catering companies that do actually make the food and businesses that want to buy it. And so the idea here is basically Uber is making this acquisition so that it can, I mean, Uber has been trying to grow not just its main business, but Uber Eats, and has been doing it, but it's also been doing both of those through a lot of acquisitions and partnerships and investments.

• 33:05Speaker 00

  And so this is one way to grow Uber Eats into a market that Uber is kind of not really in right now, which is sort of like office catering, event catering. You know, the way that Dara Khosrowshahi pitched it in the press release about the acquisition is really about, and, and I think if you read between the lines here, which is what I'm about to do, I think you have a lot of restaurants that are on Uber Eats and other services, you know, DoorDash, whatever, that don't necessarily always feel like they're getting the best, you know, sort of end of the stick, and getting like, kind of pushed around a little bit by the platforms and its incentives and the cuts that it takes, and, and sort of just how demanding it is to be in the food delivery business when you're already a low-margin restaurant.

• 33:48Speaker 00

  And so this is a way to increase potentially the amount of, you know, it gives these restaurants another avenue for where their food can go, right? Like, you might now be able to provide food for catering setups in a way that that was not previously accessible to you. So, you know, my question here is sort of like, whether or not this changes, you know, how much does this change how a restaurant thinks about its business? Does it make it something that is, you know, makes it a more viable business than it traditionally is? Because it's usually a really tough business. Or is this really going to just kind of be, you know, a little bit of extra, a little extra coin for the sort of commercial kitchens that are already out there?

• 34:31Speaker 00

  And I should say real quick, like, it's a cool, it's a cool like, exit story, I think. Like, I was looking back, we wrote about this company way back in like, 2014. They were founded in 2007 and bootstrapped for like, seven years before they even did their first raise. So to get to a point where you're doing an all-cash $2 billion exit, like, 15, 20 years later, is, is a very different story than what we see these days with a lot of the AI startups.

• 34:55Anthony Ha?

  Yeah, that's pretty much what I was going to say as well, is that it kind of, the story just kind of cast me back to like, 10 or 15 years ago, when there were a lot of different catering startups that were, again, sort of like, we're catering, but we're not a catering business, we're like matchmaking, we're a marketplace. And, you know, that excitement is not, has not sort of carried over. So I, it was a very pleasant surprise to me that not only were they still around, not only do they sell, but they sold for, like, you said, $2.3 billion.

• 35:20Rebecca Bellan?

  Uber loves buying companies. They love it.

• 35:23Speaker 00

  I mean, I, not to go too deep into this, right? But like, Dara Khosrowshahi used to run Expedia before he came in to run Uber. And I think you think about that, and you look at the way that he's oriented Uber's business on all sides of it, like, I think it just makes a lot of sense that he, he sees a lot of value in being, you know, the place to go for not just what it's known for, but for being, you know, having a lot of options from other companies and other services. And he's certainly grown the business that way and made it profitable for the first time ever, so.

• 35:53Rebecca Bellan?

  Yeah, and it's how he gets that stickiness, right? I wrote about this a few years ago, the Amazon-ification of Uber, right? Just like, feeding into the different platforms until, you know, Uber is, Uber is the thing that moves people and goods, right? Well, I guess that's a fine place to end it. We are just about out of time. Equity will be back next week. Until then, you can find Equity under the handle at EquityPod on X and Threads. Thanks for listening, and we will talk to you next time. Equity is hosted by TechCrunch senior reporters and produced by Teresa Locan-Solo with editing by Cal. Subscribe on YouTube or wherever you get your podcasts and find out what's next at techcrunch.com/events. Thanks so much for listening, and we'll talk to you next time.