Closing Bell - AI, Earnings and the Fed Drive the Market 9/3/26

Episode Date: September 3, 2026

Big Tech and AI stay at the center of the market conversation. Barclays’ Venu Krishna breaks down the latest trends in AI adoption and whether Meta is getting its strategy right. BTIG’s Janine Stichter reacts to Lululemon earnings and what the results say about the consumer. DataTrek’s Nick Colas looks at bond yields, Fed speak and the setup for the jobs report. Meantime, Kate Rooney reports on organized crime rings targeting valuable AI intellectual property. Cboe’s Mandy Xu explains what options and volatility markets are signaling about the next move for stocks. Guy Adami weighs the signals coming from gold and bitcoin as investors search for opportunity outside equities. Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

The bell is bringing an end to the trading day at the NYSC. Fans for the cure ringing the bell. And at the NASDAQ, it's tactical resources closing out the day. Welcome to closing bell overtime live from Studio B at the NASDAQ market site. I'm Mike Santoli. Melissa Leet has the day off. Stocks higher across the board as software shines and yields fall. The Dow up 600 points. The S&P 500 higher by 1%. The NASDAQ composite and the 100, both up more than 1%.

Bond yields and oil both were subdued today. The 10-year yield coming off its highs, as Fed Governor Waller indicates, he leans toward voting keeping rates unchanged at the next meeting in a couple of weeks. More on that to come. And once again, we're watching earnings. Lulu Lemon due out this hour as our Z-scaler, Samsara, and DocuSign. Let's get more on today's market gain. Sima Modi has the detail.

Mike, the NASDAQ was the standout today of 1.4%. Tech led the way with all magnificent seven names trading higher. following NVIDIA's $13 billion deal to buy, Hugging Face. Beyond NVIDIA, Broadcom led chips lower, as investors questioned the size of its quarterly earnings beat. Some of the semiconductor names like Texas Instruments, advanced micro-devices, Qualcomm, trading down about 1% on the day.

Different picture for software, which got a jolt from Snowflake's strong guide, accelerating revenue picture, Bank of America, raising its price target to $470. The stock ending higher by 16.5%. results once again easing fears around AI displacement. Check this out. The IGV software ETF up about 20% from that June 25th low, whereas the semiconductor ETF has declined during that same time period by around 16%. Elsewhere, Bitcoin bouncing higher, reclaiming that 80,000 level on Thursday, fueling gains in names like Robin Hood, Coinbase, and Circle, which was up by 16%.

Food stocks did not fare well. Campbell Soup saw profits decline and its dividend cut. Tyson lowered sales outlook due to pressure on beef prices. And this, as we count down to Tesla's cybercab event, where it's expected to unveil a new two-seater tonight. Stock is on track for its best week since early May, Matt, Mike. All right, yeah, two-seater with supposedly no steering wheel. We'll see what we get. Seema, thank you very much.

Let's get more on NVIDIA's purchase of Hugging Face. Christina Parts Nevelas joins me now to explain it all. Christina. Well, the company that powers the AI boom, InVIDIA, just paid nearly $13 billion for insurance against it. InVitya's buying Hugging Face, the open source AI platform in its second biggest deal ever. Invita's fortune today rides on a few giant customers. Frontier labs like Open AI Anthropic plus a handful of hyperscalers,

all spending on its chips to build expensive, mostly closed AI models. The risk is that the world moves on towards open models that are essentially free to download, sometimes cheaper to run, and catching up fast. many of them from China. Hugging Face is how NVIDIA wins either way. It's the home of open source AI, more than 3 million models used by 18 million developers. So if companies and governments start building their own AI instead of renting from the big labs, Nvidia now owns the place where they go to do it.

Here's how CEO Jensen Wong explained it on CNBC this morning. Our fundamental goal is just to make sure that AI advances as quickly as possible, And it's really, really important right now as the open models are really accelerating that we make sure that we provide Hugging Face the platform to continue to scale and for the resource for them to scale and extend the open model ecosystem and community. That scale is exactly where the worry starts. HuggingFace got big by staying independent and some developers already fear that Nvidia will nudge them towards its own chips and models like Nemotron. And Jensen Wong says this morning that it plans to stay open. The Hugging Face is going to stay open and neutral, and that's the promise this whole deal rests on. It's interesting.

Christine, I've heard you kind of compare this to Microsoft buying GitHub, right? So that's this other kind of meeting place for programmers, et cetera. I also wonder if from the outside, as it might serve NVIDIA's long-term finances, we think of it as potentially eventually in Vidia's App Store, meaning not so much that it's actually all third-party apps that they get a piece of. But in a way, it creates a services-based type of business that feeds off of the growth of the whole ecosystem and obviously runs on in video hardware, but is operating in kind of different growth dynamics. Yeah, so two points. So the comparison to Microsoft GitHub, it was just a few years ago acquired it. Microsoft, you know,

over $6 billion or $7.5 billion, a promise that they wouldn't, they would stay neutral and not get involved and now it's turned into almost marketing for its own AI tools. And it's grown dramatically, but it's become a Microsoft product. And so that's why I made that comparison, because it is a concern that eventually this could happen to Hugging Face. To your point with Hugging Face, they make a lot of their money through the subscription-based model and providing all the tools that developers would need, and maybe they want to get help with downloading certain models. So it is an opportunity for InVitya, obviously, to tap into that subscription-based model already. I wonder, though, how do you create app-based platform while trying to maintain neutrality?

Because you're essentially, like, will the models be catered to Nvidia compute and hardware? Right, right. And that is the major concern that they're promising won't happen because AMD, Intel, Google, all use this platform, and these are all competitor chip companies. Right. And of course, the app store, obviously, is all running on the iPhone. So it's, you know, it's closed in that sense. I'm just, I'm just, I'm, Much more thinking about it, like we're talking about revenue streams in the future and seeing how they may try to diversify away from the hardware dependence. Or, okay, or isn't it an opportunity to tap into the 200,000 companies that use HuggingFace? Right. Even smaller, bigger, sovereign, et cetera, and move away.

So, yes, this is a subscription model, but, too, another way to move into another ecosystem, much like their revenue model has changed from hyperscalers and then cloud. Remember the ACIE? Everything to everybody, ideally, I guess, is how they want to do. Do it. Christina, thanks. We have Z-Scaler earnings out. Sima Modi has the numbers. Good numbers here, Z from Z-Scaler, beat by 10 cents on its bottom line on an adjusted basis. Revenue 2 coming in above estimates at $898 million versus the $877 million estimate. Annual reoccurring revenue did jump 25% year-on-year while operating margins and gross margins were in line. Now, looking ahead, guidance looks strong above what the street was predicting for the first quarter and full year.

The report does come as cybersecurity firms were. respond to these growing cyber threats from AI agents. On that note, CEO Jay Trottery says we are uniquely equipped to help companies both combat the threats created by agentic AI and securely deploy AI agents and models. We'll look for more details on the call. Don't miss first on C&BC interview with CEO Jay Trowdry tomorrow at 10 a.m. Eastern on Squawk on the street. We are watching shares move higher here, Mike. Yeah, 12% initially, Seema. Thank you very much. We mentioned all the Mag 7 names closing in the green. Meadow with a 4% gain on. hopes for its newest AI model. Julia Borsden, joining me now to talk about that, Julia.

Yeah, that's right. Meta's over 3% gains today come on top of yesterday's 2.5% gains. That brings the stock up about 8% over the past week. Now, this comes after yesterday afternoon, meta announced the release of its latest AI model, Muse Spark 3.1, showing what the company says is meaningful progress over both his own models and its rivals. Third-party research firm artificial analysis, saying that meta outperforms Google's Flash 3.8 and Anthropics Fable 5 ahead of all the rival models, except for Claude's two most powerful models. Meta's AI, Chief Alex Wang, telling CNBC that this shows progress in coding and agentic abilities, telling us, quote, we have invested into a lot of capabilities that are going to be really

differentiating for the kinds of use cases that Mark has talked about, such as having agents that can work 24-7 on your behalf and achieve your goals and improve your life and your health and your relationships. Now, today, Bank of America with a buy rating on the stock saying the new model suggests a strong pace of AI development, noting that those AI improvements are noteworthy ahead of a launch of Meta's AI agents, which is expected soon. Mike? Yeah, been a lot of skepticism about Meta's AI efforts. Maybe that's starting to give way to some progress there, Julia. Thank you very much. Meta and every other name in the Mag 7 was hired. today, but recently those stocks have been punished due to concerns about the impact of all the

spending needed to keep up in the AI race. Should those concerns keep you away from these mega caps or evaluations come down enough? Joining me now is Vino-Krishna. He's head of U.S. Equity Strategy at Berkeley. It's good to see you. Good to see you. We did see, I guess, the sort of mag-7 type stops re-engage here. We came through the earnings season. It also strikes me that, you know, you look at a day like today where you have good reaction to snowflakes earnings, this meta-new. in video hugging phase, people are suddenly excited maybe about the consumption of AI and not just the construction of it.

Right. And just to be sure, we have been very constructive on big tech from the beginning of the year. In fact, since Q4 of 22, it hasn't worked as well this year, mainly because the concern has shifted towards the scale of their capital spending, right?

It's the majority of what we call big tech, which is max seven less Tesla. But I think what we are also seeing, that they're posting extremely good earnings, well over 30%, and the multiples are contracted about 15 to 20%. So their performance is broadly in line with SNP as of now. But going forward, that is the combination which we like, because these companies are, in the end of the day, going to dominate and lead this whole AI revolution, so to speak. So, you know, our view is that this is where you want to be.

And the bigger picture for equity investors is, you know, we also believe it's unlikely that you will get the kind of returns like you got in the last three years. So there are going to be some moderation of returns. And the single biggest phenomenon, one of your earlier speakers was also mentioning, is this notion in the equity markets, which is in a building for the last two years, is that dispersion is extremely high. If you look at correlations of stocks within S&P, within sectors, between sectors, even between these six stocks in Big Tech, they're at historic laws. In principle, it is a stock pickers market and where you pick mini themes. And those mini themes have not been easy to navigate because they change and flip in every few months. Absolutely, as they have. Just one second.

We want to get to these Lula Lemon numbers. They are out. Brandon Gomez has it for us. Right. Hey, Mike. Yeah, that's right. Take a look at shares down about 50. 15% now, I think really around this further cutting to guidance, but let's run through it.

EPS came in at $2.92 a share. Now we're not doing a comp because that does include an 86 cent tariff refund as well. Revenues was amiss, $2.42 billion as opposed to the $2.45 that we were expecting. And then comp sales, we were expecting it to be down 4.7% actually came in down 9%. And then when I talk about those guidance numbers, looks like for Q3 EPS guidance is between 93 and 98 cents a share. that's significantly less than the $2.40 that was expected. Revenue guide for Q3 coming in below expectations as well. And again, just a reminder for fiscal year guidance. There was a cut last quarter. Now we're seeing the company expecting fiscal year revenue guidance between $10.35 and $10.5 billion. That's below the $11 billion that was expected for fiscal year revenue guidance coming into the print. So again, a lot to listen to for on the earnings call. We have a new CEO starting next week. So I'll be sure to bring any headlines from that. Yeah, resetting the bar lower for sure, Brandon. Thank you very much.

Vina, just to get back to, I guess, kind of the broader market set up as we're getting into September. We've gotten, you know, the earnings in the books, came in better than anybody could have hoped. Right now we're, I guess, kind of not quite sure if the macro is going to intrude on it at all. How does that feel to you right now as, you know, yields and everything else impinging? Sure. I think the macro has definitely worsened in the last few months. So the first and foremost is the big gorilla, which is the AI narrative, that on the margin has weakened after three years of extraordinary strength. And that weakness has been driven by the sheer scale of capital spending and some uncertainty about the timeline for monetizing it, the funding issues, power constraints, the whole nine yards. The other big concern now is about rates.

Ten years are well over four and a half percent. And historically, there's been a very clear negative correlation between equities and rates once the yields, you know, touch 5%. And we would not expect that to be any different. It means that we are already under danger territory. And then, of course, don't forget that the Middle East War has rekindled and hasn't resolved. So I think, to your point, the macroconcerns are bigger. And on top of that, this is an election year. And if you look at the history of election years, typically S&P does worse off in an election year.

than a non-election year, midterms. Yes. And then the worst returns are in August and September. Yeah. So I do think that you can brace yourself for some choppiness. And there's still more upside, not a lot. But I think we remain incrementally positive. And then just really quickly, do you have the sense that investors collectively

are braced for a little bit of that choppiness? Because the talk is there. I just don't know if the positioning follows. Well, interesting. Yeah. And the positioning side, the retail front is already cautious. So retail was cautious for a year. They jumped back in after the July sell-off, but once again, they'll step back.

The excessive enthusiasm is coming from the institutional site. And they gleaned that information from flows and also the derivatives market. But if you look at, we have a euphoria index, which mainly tracks retail sentiment. That actually is high, but it's moderated a fair bit from peak levels. So the question is, what's the source of the optimism? part of it is the fear of missing out, the former trade, and it's coming for the institutional side. Right. Interesting. We'll certainly keep watching it, Vina. Thank you so much, Gita Krishna. Sticking with Lulu now, after earnings crossed just moments ago, a pretty adverse market reaction with me now to parse through those numbers is, and what it all means is Janine Stichter.

She is BTIG Consumer Retail and Lifestyle Brands Analyst. She got a neutral rating on Lulu. So it was a lot to prove here, Janine. It doesn't seem like this quarter. made any progress toward doing that? No, and I think the really interesting thing to come out of this, we'll have to see what they say on the call. But if you break down the results, which were still digesting, Erica, which has been the source of the issue, was kind of in line. It was down high, single digits.

They had said it would be down low double, so if anything, a little bit better. But the real surprise here was China, which was actually down. They had guided to strong double-digit growth in the quarter. And I think what happened there, we had done some survey work back in July around all of these headline issues they had called out in Q1, whether it was Chip Wilson or the Forever Chemicals controversy. And then after that, there was this additional controversy with the Great Wall of China. And it seems like, based on our survey work, and now we're seeing it in these results, that China really materially slowed in the quarter. And that's a problem that they haven't

really had to deal with before. Up until now, the issues have really been concentrated in the U.S., in the Americas. And this is an entirely new issue that seems to be popping up following the Q1 results. I mean, it's clearly, you know, a pretty tough series of tasks for an incoming CEO. How does, how would you prioritize attacking these issues? I think right now what we're going to have to see happen and look, Heidi's start, Heleneal, we've been waiting for her to start all summer. She starts next week. Whereas she'd assumed taking the role, assuming that this would be a task of driving Lou Lemon's next stage of growth, I think from here it needs to be taking a step back.

Priority number one has been and should remain correcting the America's business. and that's everything from marketing to product to improving brand image. But now she's going to have to deal with a whole other set of challenges in China. I wouldn't be surprised if that means taking a step back from the unit growth that they've been driving in that market for the time being while they figure out what's going on. Because right now they need to focus mostly on the Americas, get that right, and then they can figure out what's going on China. I mean, the stock obviously is, you know, numbers are probably going to keep going down, obviously. But before this report, it looked like it was cheaper than it had been. ever kind of trading, I don't know, on part with like VF Corp. Is it your take that it doesn't

yet account for the struggle that might be ahead? Yeah, I think the challenge and why it appeared so optically inexpensive is that we didn't really know where the earnings floor was. And there was definitely an assumption that they would have to take numbers down. I think you saw that reflected in consensus. But this is when you look at the number they're guiding to, we're looking at, you know, below $9 now stripping out the 86-10 tariff benefit, something in the $8 range. And we haven't had to deal with all regions decelerating. at the same time before. So this just feels like a much bigger fix than what we thought two or three months ago, certainly more than we thought six and nine months ago. And I'd say

at this point we don't have good visibility if that's even the right number. Now we have a new CEO stepping in. She may choose to step back from certain growth initiatives or certain categories. And that might be the right move long term, but it might mean that certain areas that have been driving growth, whether it's new categories or new geographies, we take a step back from focusing on expanding those for the time being. And that adds of another leg to the earnings. decline. And the other area that we talked a lot about is marketing where I think there's a view out there that they're not putting enough money behind marketing. And now with everything going on in terms of brand perception, you definitely might need to see her step that up over the next few

years just to sustain the revenue growth that they've had. Right, which would restrain earnings power, I suppose, for that period as well. Stock down almost 15%. Pretty making a new low below where it was a couple of months ago. See how it goes from here during the call. Janine. Stichter, thanks so much for joining. Thanks for having you. All right, today's stock rally getting an assist from the bond market yields falling after Fed Governor Waller indicated he could vote to keep rates unchanged at the next meeting. But there's still two weeks of data before then, including a jobs report tomorrow. We'll discuss what the Fed's facing and how they're likely to handle it with Nick Coles.

Next on opening. Here, breaking news on Adobe, Sima Modi has the detail, Sima. Mike, Adobe now has a new CEO. The board of directors today announcing that Anil Chukharvati, who is the president, The president of Adobe's customer experience orchestration business will be the CEO of Adobe. He will become president and chief executive officers starting on December 1st. Shantanu Narayan, who announced he's stepping down as CEO back in March. It was March 12th will be the company's new executive chair.

This search for a new leadership of Adobe took many months. And interesting now that the company has gone with an internal hire here, someone who has been with the company for a significant amount of time and focuses on customer experience, Mike. We're watching stock down about 2% here in after hours. Interesting. Seema, thank you. Well, another company filing to go public, Brandon Gomez, has that story.

Brandon. Hey there, Mike. Yeah, we heard about this late last month, but ORA is going public announcing. Today it has filed for an IPO and plans to list on the NASDAQ under the ticker OURA, ORA. The company reported $907. million in revenue for fiscal 2025,

according to the S-1. Giving investors their first detailed look at the business, ORA has grown beyond its roots as a sleep tracker expanding into heart health, women's health, metabolic health. It also launched its AI-powered ORA advisor and has built out retail and distribution partnerships. The IPO is being led by Goldman Sachs, Morgan, and J.P. Morgan. Now, ORA was last valued at about $11 billion last year after raising more than $900 million. The big question for investors is going to be, will Wall Street see this as just another wearable company or possibly more, Mike? Yeah, gadget companies have sometimes had a tough road.

We'll see. How goes, Brandon, thank you very much. Let's take a look at bonds here. Yields taking a bit of a breather today, helping to boost stocks. That's after some dubbish comments from Fed Governor Christopher Waller at a speech in Washington, saying he supports leaving rates unchanged if inflation continues to make progress. The latest from CME Fedwatch showing a coin flip in terms of how traders are betting the Fed will act later this month, the hold camp getting a big boost in the last day or so. Joining me now is Nicholas Colis. He is Dat Trek research co-founder. Nick, good to see you.

Good to see you. Somewhat unusual to be, you know, less than two weeks out from a Fed decision. We're still in coin flip mode. Obviously, it would be a change of direction as well. The last move last year was a cut. How do you feel it's shaping up and how does it interact with the data we're going to get? Well, I mean, welcome to a Warsh Fed. Yeah.

One doesn't chat very much. Weller's comments were, I think, a bit unusual. I've actually given a little bit of forward guidance, at least from one voter. but this is the kind of market we should expect. If you get the 70% odds on a rate move ahead of a meeting, that probably is as close to certain as you're going to get for the next four years at least. What is your, I guess, read on what exactly has been moving longer-term yields higher?

So clearly we've repriced the Fed to some degree, the Fed path from here. Oil up. Obviously, there's some of this stuff about fiscal concerns, demand for debt. I mean, all of it seems to be in the mix. How much of it sort of matters? much of it holds water? So we were up like 62 basis points or so on tens for the year and at pretty high levels, uncomfortably close to 5%. The vast majority of that move is real rates. It's not inflation expectations. It is just the market saying, you know what, we've had 15 years and

two months of recession. This economy is extraordinarily strong and the neutral rate of interest is probably higher. We had an oil shock this year that in 1990, the similar thing drove us into recession almost immediately. And we haven't seen that this year. So between the trade shock last year, the oil shock this year, other uncertainties, and still no recession, the market's very rationally saying, you know what, neutral rates probably higher than 3.63, where we are today on Fed funds, and that's what the 10 years repricing. And so, you know, obviously we get the non-farm payrolls tomorrow. It seems like that has less potency, right?

The labor market is not really the first place we're looking to figure out the next move for the Fed. It's kind of been give or take zero. Wage inflation is not an issue. CPI next week, I mean, it's not the Fed's. target number, but I suppose that's the last chance to sort of steer the view. Yes, it is. Absolutely. I mean, Warsh was very clear at Jackson Hole saying he's not worried at the labor market, and he's right not to worry. But let's take a step back. It's also kind of

a miracle that we don't have to worry about it, just given where oil prices have been. Yeah. So let's not discount tomorrow as saying, oh, it's just a zero job report, whatever. It's important. It reconfirms the idea the economy is still okay. Yeah. And okay is a way to put it, certainly from the perspective of jobs. You know, we did get the Jolt's data this week. Again, sort of like this low energy equilibrium it feels like is where we are, right? Low higher, low fire, not a huge jump in, in fact, a decline in openings. So, you know, is that giving us much information? Yeah, we do this thing called the take this job and shove it indicator, which is quits is a percentage of total separations.

Basically, out of 10 people who left their job, what percentage did it because they voluntarily chose to leave? And it's running around 60% right now. 60% of all separations are quits, which is a kind of strong. than average reading, which says that, you know, quitting is a function of opportunity and motive. People are still at the margin quitting a little bit more than historically average, which says the underlying tenor of the market is pretty strong for labor and reconfirms Warsh's point of view. Yeah, I think we do have the demographic situation where you have so many retirements and there's some attrition there. I mean, companies maybe don't have to actually

let people go as actively as they did before. Yeah, and I'll tell you, it took so many companies so long to hire in 22 and 23. And it's taken so long. to train them that you don't want to let them go unless you absolutely have to. You mentioned that yields getting to what feel like high levels. If they obviously went high in a rapid fashion from here, maybe it does impinge on equity evaluations and things like that. But in absolute terms, you look at a long-term chart, it's like, what are we getting excited about? You know, four and three-quarters on the 10-year when, you know, in the late 90s, we're at five and six percent, and we felt like that was low rates because of where they had come from. Yeah, it's all my life is

always about baselines, right? We baselines in the 2010s at ever lower rates. We hit 1% on tens in 2020 and 2021 and now we're at 5. So it feels junky. It feels heavy. But the market's dealing with it's okay because it comes from a genuinely good place, which is expectations for strong economic growth. You can have a 20 PES&P and a 5% 10 year coexist very happily if they're both anchored on strong real forward growth. Yeah. And I guess strong real forward growth, plus you have, you know, commodity indexes are rallying and things like that. It just feels like a higher metabolism outlook, whether it's going to happen or not, would imply over cycles more a tighter fed than it would otherwise be the case. Absolutely. That's the 10 years saying exactly. Yeah. All right. We'll see up.

Things break tomorrow. And next week, Nick, thanks so much. Thank you. Good to talk to you. Up next. A CBC exclusive about how foreign cybercriminals are stealing data from the dark web and then using that information to hack into Americans' AI accounts like Claude, Chat, GPT, and Gemina. Google OpenAI and Anthropic have poured many billions into building the world's most advanced AI platforms, but there's a new threat to those investments. Experts say foreign companies are trying to illegally access those systems and use them to train their own cheaper AI technology.

For the first time, Anthropic is speaking on camera about what these attacks could mean for its business and the national security risks. According to the AI lab, the attackers are doing this by ripping off American technology. Kate Rooney has the exclusive. Cybersecurity experts say foreign criminals are getting stolen credit card information and AI accounts from the dark web to illegally access platforms like Claude, Gemini, and ChatCTPT in places where these platforms restrict access, like Iran, China, North Korea, and Russia. But it's not just so they can check out the latest technology. Experts tell CNBC they're trying to steal it. It's a hard-to-trace tactic called distillation.

To do this, hackers will ask AI models thousands of questions, then they use those answers to train their own smaller AI platforms, which are often offered at a fraction of the price. Travis Lanham is the chief technology officer of cybersecurity firm Armaden. He says not only are cyber criminals using individual consumer accounts, but larger business accounts that allowed them to ask many more questions. and allow them to execute this at scale. Those types of accounts that can be used to ask a lot of questions or do a lot of prompts and then capture that data

versus the smaller limits that are applied to the individual user accounts. Why isn't that stand out to an AI company? I would think something like that would be easy to spot. These companies are serving billions of requests, and the millions are relatively small compared to everything, and it's just sneaking in and trying to look like the rest of the crowd. Jacob Klein is in charge of shutting down the global cyber criminals targeting Anthropic. He says beyond the risk of IP theft, Anthropics' concern is that these stolen models can be misused.

These unsafeguarded models can pose a national security concern. They can be used for things like hacking into critical infrastructure, hacking into governments, into companies. Have you seen any evidence of distillation attacks resulting in cybersecurity, issues or anything on that national security level. We saw a campaign from a Chinese-based entity conduct espionage at scale in an automated fashion leveraging our technology. We disrupted this, we improved our safeguards, but we see this on an ongoing basis, bad actors trying to use Claude to conduct malicious cyber attacks.

Anthropic has publicly alleged its chatbot. Claude has been distilled by three Chinese labs, moonshot, Minimax and Deepseek, and says these labs generated over 16 million exchanges with Claude through approximately 24,000 fraudulent accounts. It exclusively told CNBC that Moonshot recently developed a new model that distilled off of the latest Claude model. DeepSeek, Minimax, and Moonshot did not respond to CNBC's request for comment. And Mike, Washington is very much paying attention. In a memo from April of this year, the Trump administration wrote the distillation that undermines American research and proprietary information is, unacceptable and that it would explore a range of measures to hold foreign actors accountable,

Mike. Yeah, Kate, you asked, I think the key question about, you know, exactly how these attackers managed to do this without being in real time detected. Once those sort of models are built or their distillation has resulted in some kind of a product out there, are we able to identify where it came from? In other words, can these labs basically say, yes, that's our stuff there? That is what Anthropics said. And it was interesting when we started reporting out this piece, Mike, we were asking experts,

and they really said you have to go directly to Anthropic or Open AI or Google, because they are the only ones who can really tell that this has indeed been distilled. If it's an open source model, sometimes there's hints or clues, but the lab said explicitly, or at least Anthropic said explicitly, moonshot, which was the one that got so much buzz. And at the same time, is widely adopted in corporate America. They talk about the risks of not having safeguards or what they can be used for on the downside. On the upside for businesses, they're a lot cheaper.

They're being used widely. So it is really a rock and a hard place, I would think, for the Trump administration especially, and for regulators to say people really like this open source technology. But at the same time, some of the most popular versions here are versions that have allegedly been stolen from American IP. And they are really faced with this issue right now. Most of the companies we spoke to said it's almost impossible to completely stop without banning open source, which most CEOs would disagree with and say that's not the right move because you don't want to shut down competition. Yeah, I mean, Napster was pretty popular.

I mean, you knew where the songs came from, but at some point there had to be, you know, kind of a line drawn. We'll see. Kate, thank you so much. Thanks, Mike. Time for a CNBC News Update with Brandon Gomez. Hey, Mike. Another twist in the murder trial of Lindsay Clancy. Her lawyer asked the judge to dismiss a juror who he says, was causing the group to be deadlocked.

He declined and ordered the jury to keep deliberating after reminding them one by one about the rules for reasonable doubt. The jury has already told the judge twice in six days of deliberations that they can't agree whether Clancy should be held criminally responsible for killing her three children. The FBI announced the arrest today of a podcaster in Los Angeles who was accused of threatening President Trump's life. According to federal prosecutors, Benjamin Southworth

allegedly posted videos calling for the president's assassination. They say he also published the home address and other personal information of the president's lawyer, Jay Sekulow. And meta CEO Mark Zuckerberg reportedly called the potential establishment of a national AI regulator flawed. According to Politico, Zuckerberg said he opposed the idea in a phone call with President Trump last month. The proposal would set up an independent organization modeled on the financial industry regulatory authority or FINRA. Mike? All right, Brandon, thank you. Well, the volatility index, near a one-year low.

But up next, we'll discuss whether the options markets are signaling a market shock on the horizon. Welcome back to closing bell overtime live from the NASDAQ market site. Big gains for stocks today. The Dow up 624 points, 1% gain for the S&P 500, the NASDAQ composite, NASDAQ 100 as well. Small cap rustle higher as well, but not as much as the bigger stocks. Software stocks were big gainers following snowflakes numbers. That's not gaining 16% on the day. But a couple of software names getting hit,

after hours, guidewire software, off by 19, almost 20%, fourth quarter earnings of 99 cents a share compared to the estimate of 94 revenue of 411 million. That was a slim beat, but the first quarter revenue guidance seems to be the issue. 378 million at the high end of the range. The consensus forecast is 387. Asana, beating by a penny on earnings and a couple million on revenue for the next quarter, though, guiding in line with the estimate on revenue and a penny below on earnings, and that stop is falling more than 12%. It's been a busy summer for the markets, lots of news, earnings, and a move higher in yields. But through it all, the S&P 500 has stayed fairly steady within a couple of percent of record highs.

Along with that, the SBO S&P 500 volatility index the VIX has also been relatively calm, falling below 15. But should we be prepared for an increase in volatility ahead? Joining me now is Sibo's head of derivatives market intelligence, Mandy Shoe. Mandy, good to see you. Good to be here. Yeah, equity volatility subdued. We've had a lot of kind of internal rotation, kept the S&P 500, pretty much calm for the most part. On the other hand, bond market volatility has perked up. So how does that play into the picture? Yeah, so I think you bring up a really notable divergence that people are pointing out right now. We're seeing increasing volatility in the bond market. Obviously, yields going higher on the back of the hawk shift from the Fed. At the same time, if you look at the VIX, like you just mentioned, VIX sub 15 is very low.

So we've been getting a lot of questions around, is this a sign of complacency in the market? And here's how I would, you know, frame it. The VIX at its core is a measure of short-term equity market volatility, right? It's expected 30-day volatility of the S&P 500. So a lot of the risks we're talking about with, you know, in terms of inflation, path of rates, those are longer-term risks. So where that would be reflected in the options market is longer-dated options. So if we look at, for example, S&P one-year implied volatility, that's relatively elevated compared to the VIX.

The spread between the two actually near a one-year high. So that's the options market saying, you know, while things are calm right now, if you look at the VIX, you know, expect more volatility ahead. Well, you have to account for the possibility that were somewhere closer to the lows in volatility than the highs. This other dynamic's been fascinating is I mentioned the rotation, the fact that you had for a while a lot of single stock jumpiness, a lot of volatility, and yet it was kind of being offset inside the index and the overall index was fine. So that has come in a little bit, right? I mean, in terms of single stock movement.

Exactly. Yeah. So one of the things we've noted over the past couple months is just the incredibly high volatility at the stock level that wasn't being reflected in the index level precisely for the reason you mentioned, the record high dispersion, low correlation in the market. I would say this past earning season has been really pivotal in addressing some of the fears underlying that bid to single stock volatility, namely the fears over the AI trade. That was primarily what was driving the bid to single stock volatility.

And this earnings season has been, I would say, largely very constructive and helping calm some of the fears around kind of potential pullback and spending from the hypers, you know, the gap between spending versus revenue, et cetera. So since then, we've seen a very significant collapse, I would say, in the premium in single stock volatility relative to index. So the two indices that I look at VIX EQE, which is our measure of kind of average single stock volatility in the market versus the VIX, which is the VIX, which is a measure of kind of average single stock volatility in the market, versus the VIX, which is a VIX, which is a cost, which is, which is, which is, is the benchmark for the S&P index. If you look at the gap between that two, it's coming about 15 points. It went from a record high of 35 to now about sub 20. That's a very big move, the biggest we've seen in over a year. And that tells you people, a lot of the fears have shifted away from the micro towards more of the macro. And yet, it's fascinating how you set that up because when we were at that record widespread between single stock volatility

and index volatility, it actually was being presented as a potentially bearish setup, right? Because it was like, well, once correlations rise and all stocks start moving more together, that's often when you get a little bit of a market accident has not really happened. No, so I would say the average single-stop volatility has come in. Correlations hasn't really been bit up. Ah. So that's, so it hasn't happened yet. So there's a lot of concerns right now about the longer-term outlook, but so far right now,

correlation levels are still near historic low. So when that happens, certainly that would be, you know, that could drive index volatility, VIX a lot higher. We still haven't seen that. just the risk around single stock fundamentals have come down a lot. So single stock volatility lower, but correlation levels really haven't changed. That makes sense. And then just quickly on the bond volatility where we started, I mean, it's often mentioned that, oh, higher yields are not a problem unless it's a very volatile move.

Fixed income volatility has a way of spreading. Are we at critical levels in that sense yet? So, yeah, I don't necessarily put stock in a particular level. I don't often throw around a level, but I would know that right now we've been in a persistently a positive equity bond correlation environment. And that is really impactful for people's portfolios, right? Because that means fixed income is no longer diversifying your equity risk, which is a primary reason or one of the main reasons people hold fixed income in the first place.

So what we've seen in this shifting environment is more and more people look to, for example, options to give that more specific protection or that more specific outcome that they need. So that's something that I would point out to is more important than a specific level of yield. is that correlation set up. Makes sense. Mandy, great to see you. Thank you. Great to be here.

All right, Bitcoin has been bouncing back, rallying nearly 30% over the past month. Up next, Fast Money's Guy Adami on how he's trading the cryptocurrency and crypto-related stocks. Closing bill over time. We'll be right down. For the past month, Bitcoin and gold have been outpacing the S&P 500 with double-digit gains. But broaden that out to the start of the year and stocks have still been a much better bet. Could the recent gains signal a change in that trend for Bitcoin and gold?

now is Risk Reversal Media, co-founder, Guy Adami, also a fast money trader guy. Hello, Michael. Pretty strong bounces. I think they're almost on the verge of, is this a counter-trend move? We popped above the 200-day average for both Bitcoin and gold, or is it, you know, the start of something? Is it the start of something? So I would take, it's the start of something. I'm not suggesting I'm right, but you study this stuff and you pay a lot of attention to a lot of things. 2013, seemingly out of nowhere, the Bundesbank announced they were repatriating their gold. By the way, gold that had probably been stored elsewhere for 100 or so years. And you say,

well, what's the big deal? They obviously saw something. Fast forward to yesterday or two days ago, all of a sudden, the Netherlands out of nowhere, 86 tons of gold, they're repatriating as well. So you say to yourself, okay, central banks are sort of onto the game that, yeah, we may own the gold, but unless we have it in our borders, we might not own it. So you're starting to see that, and I think that speaks to, this is just me, this whole de-dollarization thing that you talk about and we've heard about. So to answer your question, yeah, I think it's something more than just a counter-trend move. Yeah. I mean, de-dollarization is one of those things where a lot of people around the world probably feel like it'd be desirable. On paper, we'd love to do it. Not as easy to do.

No, it's not. I mean, obviously, through other currencies in particular, we did, of course, see this dramatic move with the yen overnight. So big yen rally dollar down. And by the way, it happened yesterday. I know you know this as well. Yesterday, about 9, 30, 10 o'clock, there was a moved higher in the yen, lower in the dollar. I tweeted something out because it came out of nowhere, and again, it happened today. So remember, this is ahead of a long weekend. Obviously, Treasury has a vested interest for the end of strength, and as does a bank of Japan. We're in September.

B.OJ meeting, probably going to raise rates. And who knows what's going to come out over the next 24 to 48 hours in terms of Operation Swiss and those things. But with all that said, getting back to the original conversation, all roads lead to gold. And I'll throw this out there because why not? The United States marks their gold at $42.22. We own, I think, 291 million ounces of gold or something like that. Back of the envelope, if we were to mark to market, that gets you to a trillion. I guarantee you, and I think you would agree with this, not that it's going to happen,

but there are conversations to which what happens if we mark to market our gold position? Can we do that? And what are the ramifications if we do? I'm not suggesting it will happen, and I'm not one of these, you know, getting away from the dollar as the global currency. But I know those conversations are being held. Yeah, I do find that thought interesting. I know it's floated out there once. I mean, look, the world knows how much gold there is. We know what the current price of it is. If the U.S., you know, for accounting purposes marks it up, does that help at all? I don't know. Does it show desperation to say, listen, credit rating agencies,

creditors of the world, we're okay for the money? No, exactly, which is why I think, you know, pros and cons of that. Again, I'm not saying it's happened. It's something to think about over the weekend, But back to the original question, yeah, I think gold was higher. We'll keep an eye on it. Guy, thank you. Thank you. In a bet. Up next, another check on our after-hours earnings action and the data that could move the markets tomorrow. Closing bell overtime live from the NASDAQ market site.

We'll be right back. Let's get one more check on today's earnings movers. Lulu Lemon down after missing on revenue and comparable sales, down 9% worse than expected, taking a big hit to next quarter's earnings as well, seeing 93 to 90s. The current consensus is $2.40. Shares down 18.5%. Semsara and after hours winner, earnings of 20 cents a share. The estimate was 16 cents. Revenue also better than expected. Its third quarter guidance seems slightly ahead of current estimates. The stock up 13% right now. Tomorrow and overtime will be joined by Samara's CEO to discuss those numbers. Let's get you set up with tomorrow's trade today. It's the August jobs report could be a major market mover. Economists expecting non- farm payrolls to rise by 53,000 unemployment rate to hold steady at 4.1%. Average hourly earnings are seen increasing by 3% year over year. That is going to do it for overtime today. Fast money begins right after this break.