November 6, 2026
WHEN: Today, Friday, November 6, 2026
WHERE: CNBC’s “Squawk on the Street”
Following is the unofficial transcript of a CNBC exclusive interview with OpenAI CFO Sarah Friar on CNBC’s “Squawk on the Street” (M-F, 9AM-12PM ET) today, Friday, November 6, live from Pebble Beach, CA. Following is video of the interview on CNBC.com: https://www.cnbc.com/video/2026/11/06/openai-cfo-sarah-friar-the-demand-for-ai-is-real.html.
All references must be sourced to CNBC.
SARA EISEN: Welcome back to “Squawk on the Street.” Voters sent a clear message in this year's midterm elections in favor of a pause on data center construction in states like Texas, Ohio, South Carolina, and Georgia. Democrats who campaigned on a data center pause not only won their races, but flipped seats held by Republicans. The nationwide rebuke causing investors to rethink financials behind the AI build-out. And joining me here now in Pebble Beach for a CNBC exclusive is OpenAI CFO Sarah Friar. It's great to see you again.
SARAH FRIAR: Sara, always great to be with you.
EISEN: And, boy, what timing. What did you make of the of the message from the midterms on AI and data centers?
FRIAR: Yes, I always take a step back first and say, look, the demand for AI is real. I heard Jane Fraser on here earlier with you say the same thing. What does that mean? At OpenAI, we're building a durable, strong business. My job is to manage through scenarios as they come up. We saw a breakout Q3. Our business grew 70 percent quarter over quarter. Our enterprise business doubled. So what's that telling you? It's saying people are actually using AI and getting value from it; 1.2 billion people across the globe come to us every single week to get things done, 300 million for health and wellness, 200 million people for personal finance. And, in business, it goes from small businesses, who now have the best CMO in the world at their fingertips, all the way up to large corporations like J.P. Morgan Chase, which was one of our customers who joined us in Q3, particularly focused in areas like cyber. So, when I think about the midterms, I need to keep thinking about, how do we keep investing to meet that customer demand? It starts early with having built relationships with the, all of the hyperscalers. And it will continue to mean that we raise capital. We've done it at unprecedented levels. But it also means we're going to keep building out those relationships with communities so we can get the compute we need.
EISEN: I was going to say, all of this rests on building out critical infrastructure that now Americans hate. How do you do that?
FRIAR: Well, first of all, you have to go meet communities where they are. You have to listen, learn. You have to set a plan and then be accountable. So, when we go into communities, what do we hear? First of all, there's concerns about things like electricity. And we, OpenAI, have promised we will never raise your electricity bill. In fact, often working with utilities, we can help invest in that grid. Number two is water. When you look at a large data center footprint, it's usually about two Olympic swimming pool size worth of water, which we truck in. And then it's like a car radiator, right? It doesn't really get lost. In fact, most of the water is for the people working in those data centers, and those are good jobs. And then I think the third thing is worried about noise. I was just in Georgia two weeks ago talking to the community about how we would build a 300-foot-high berm around a data center to be able to control noise. But what else can we do in those communities? In Georgia, for example, we've invested in free school lunches. I actually went and ate school lunch. It was, it's actually delicious. We're seeing 70 percent lift in kids eating lunches and breakfasts. That's a great thing to do for a community. We're building a park. We're doing an AI summit for small businesses. So I think what we need to do—
EISEN: Don't you think you should have done all this before the election? I mean, you came out with this new policy yesterday that feels very reactive, when it feels like it's too late.
FRIAR: Well, I think we have been very consistent throughout on, here's how we will show up, and newer policy, but you can go back a year or two years from now and see how OpenAI is showing up with communities. And I think, when you go in at a grassroots level, you get a very different feedback loop than when you just listen to a more binary outcome, right, here on CNBC, for example.
EISEN: So, why do you think it's not working? Do you think that people are fundamentally against the data centers in their communities or against AI and worried that it's going to steal their jobs and potentially extinguish humanity?
FRIAR: Well, I think the jobs topic is a good one, because you saw the jobs report out this morning. So, productivity is up, unemployment rate is down. In fact, wages are up as well. And that's been very consistent with, again, what our economic research has shown. So if you look at a person's job on ChatGPT, their jobs are expanding. What does that mean? About 40 percent of what they come to ChatGPT for goes far beyond the remit of what their job description might say, right? In my finance team, 90 percent of the tasks they do are agentic in nature. They're coding. They just don't know it because they're building. They're doing marketing. I'm always happy when finance people are being thoughtful about how to market things. And then, if I look at what does that mean for a company like Cisco. For Cisco, we deployed our technology. We're seeing their road map get put together 20 percent faster, and they're literally saving thousands of hours per month on areas like coding. So AI is absolutely having an impact on people's lives, on their jobs, but in a very empowering, enabling way.
EISEN: I mean, you mentioned that enterprise is a big new source of growth. I was going to ask, but how much of the business is it and where does that go?
FRIAR: Yes, so in Q3, as I said, our enterprise business doubled quarter over quarter. So we're now a very balanced business, 50 percent from consumer, 50 percent from enterprises. What's driving that? Number one, coding was an early home run, I think, for the whole AI ecosystem, but we're starting to see areas like cyber really heat up. We launched GPT-6 Cyber yesterday. Look, this is an area where I think every company in the world, no matter how large or small you are, needs to sit up and take notice. Chinese models today are getting very cyber-capable, and so we want to make sure we're putting the best tools in the world into the hands of the defenders. What does that actually mean in real life? In August, we found two fairly major flaws in PHP, which is the underpinning to the world's Internet.
EISEN: Yes.
FRIAR: We worked with the defenders to fix those flaws. So, by the time they were discovered or announced, we had fixed those vulnerabilities. That means 70 percent of the Internet is now safer, and we didn't see the sort of outcomes that could have happened, so a lot going on in enterprises on the cyber side. And then finally, just ChatGPT Work, that's our agentic tool. We rolled out Dots at DevDay.
EISEN: Yes.
FRIAR: Dots is a proactive virtual assistant for every single person in your organization. Uber started with 1,000 employees. Today, they're rolling it out to 30,000 employees. So I think corporations are walking the talk, not just talking the talk.
EISEN: I was going to ask. So that's more corporate-focused, because how it stacks up to things like Meta's Muse, which also had a really strong debut.
FRIAR: Yes. So, when you look at something like the agentic proactive virtual assistant, we started in the enterprise for a couple of reasons. Number one, we think those are some of the hardest tasks, where you can see most value created, and so it does allow for, honestly, a better business model for us to start. It also is, I call it training at altitude. If we get it right for enterprises, where the security bar is high, that means we can take it to our 1.2 billion consumers. And I think the third thing is, look, we live in a more compute-constrained world, so we have to start in the place where we can maximize the outcome for a durable, strong business. But, of course, our mission is AGI for the benefit of humanity. So we ultimately want to bring it through our entire base, so that everyone in the world gets to experience what it's like when an agent can wake you up in the morning, can tell you, here's what you should get going with today. It might be buying a birthday present for my husband next weekend, or this weekend, actually. Or it might be—
EISEN: Don't mix that up.
FRIAR: Helping me respond to a very intricate finance question that's come in from my compute team.
EISEN: But it also might create cyber risks. I know you mentioned that's a growing book of the business, but it's also a growing concern after the Hugging Face episode that we had and this idea that AI is making the cyber world a lot more dangerous.
FRIAR: Yes, so security and alignment and safety has to start within OpenAI. And so let's look at what we have done there. Number one, we have paced the frontier when we have needed to. So we have actually held back models that we felt were at a point where we needed to get safety and alignment stronger.
EISEN: Is that still happening?
FRIAR: Yes, so for example, the model that actually solved the Navier-Stokes, that Millennium math problem, which is a little nerdy, but could do things like help with artificial heart fluid dynamics, we are still holding that model. We have not yet released it. And that's an example where we needed to do more work internally to really make sure we feel confident and the folks, the regulators that we work with feel confident that that is a safe model. At the same time, we want to make sure we get other tools out into the world. GPT Cyber yesterday is a great example of that. So I think, absolutely, concerns are legitimate. And it comes back to listening to make sure we do the work internally, but then we're also getting these tools into people's hands.
EISEN: You mentioned fundraising, that that was part of your remit and you're working on. Where are you on the fundraising efforts? Because reports indicate that you are currently fundraising, but at a lower valuation and raising less than originally expected. Is that true?
FRIAR: No, we are in a current round right now. We always want to be serendipitous when we can. A good CFO always raises money when she can. We will raise at a $1.25 billion pre-money valuation. The round is ongoing, but we feel quite confident in it.
EISEN: But that was originally supposed to be $1.4 trillion last month.
FRIAR: I think the original, like, started with news from another outlet talking about 1.2. In the end, 1.25, remember, is almost a 50 percent lift in valuation from the round we closed in March. So I genuinely think that that is a good indication of our Q3 momentum, and, in particular, how that momentum carries into 2027.
EISEN: Yes, I think Greg Brockman, the co-founder, said yesterday that it was a rough start to the year, but an inflection in Q3. What are you hearing from investors right now, especially just in the last few, October was a rough month for the public AI trade, and certainly yesterday was pretty ugly. So what do you hear from investors?
FRIAR: I think investors were really impressed with the Q3 that we put up, right? Not many companies grow their business 70 percent quarter over quarter at the scale that we're talking about. The fact that our enterprise business doubled, I think that's really kind of shown what we have said all along, is, we're building one unified platform. When we get a breakthrough in a frontier research, it can help the consumer, it can help the small business, it can help the largest corporation in the world, it can help governments. That is what differentiates OpenAI. I think investors like where we're strong and different. Let's take our ad model, for example. So, when you have a consumer business that's as big as ours, we're almost 20 percent of the queries of Google right now. That's unlocking a whole new total addressable market for us. So we launched our first ads in the early part of '26. By August, we had a billion-dollar run rate business on our hands under 200 days, fastest ever. Why?
EISEN: Yes, it's like Meta and Google had a baby, and that's your ad business.
FRIAR: Exactly. So we have this high consideration, right? You come to ChatGPT to get something done. And we have high personal context. So we can get something very particular for you. So advertisers are great, getting great outcomes. Consumers are getting great outcomes, and you see it showing up in our business. So, next year, we will have this tailwind of a brand-new business in a very large addressable market with a very high gross margin.
EISEN: Does Anthropic going first, IPO, present a complication in terms of raising money right now for you?
FRIAR: Look, I always say, raise money when you can, but also run your own race. And so, for OpenAI, we need to come back to what makes us different. So we are a much broader-based platform. We serve consumers, 1.2 billion of them, and enterprises. We are broader in our frontier research, so we invest not just in LLMs, but we have the frontier model in images, in voice, right? The fact that AI is creating this whole new input mechanism, I talk much faster than I type, and I definitely code in English much faster than I code in any other language. I think the areas that we have excelled in—
EISEN: This is what you think your competitive advantage is?
FRIAR: Yes, absolutely. And we have also invested in compute early. I think we were prescient on that last year. We talked about it at that time. And it includes investing in things like our first-party chip program. We announced our chip in the summer. I think people were blown away by the performance. We think that's going to continue to let us serve our customers at a much lower cost. That's a competitive advantage, but, importantly, it's how we will achieve our mission.
EISEN: Did you look at the S-1, the Anthropic S-1?
FRIAR: Of course I looked at the S-1.
EISEN: Did anything surprise you?
FRIAR: I think on a couple of fronts for our business. I mean, number one, it kind of re-underscored just the breadth of what we're doing. When you start thinking about a consumer business alongside an enterprise business, that is uniquely powerful. And then I think the second thing is how we continue to work with partners. We just launched with AWS in the summer. We made the announcement about GCP as a partner just last week. And so, again, these should be quite incredible tailwinds for us as we move into '27, because we know that customers, we want to let them buy where they want to buy. If you want to come direct to us, we're here. If you want to work with a partner, we're here for you. We understand that that's how enterprises, typically what their motion is. And so I'm excited to see us now show up that way.
EISEN: Eighty-five percent gross margins, 45 percent compute margins. Are your numbers similar to that?
FRIAR: We have not talked publicly about our margins.
EISEN: I know.
FRIAR: But we do continue to invest in our business, because, again, what we see is the ability to build these long-term embedded options in the business, so not just the business today, which we said doubled quarter over quarter, but also it allows us to put in place businesses that will grow over the long term. And, to me, from my prior world as an equity analyst, when I think about being a CEO and a CFO previously, investors like to know that your business is strong today, which ours is, but they also want to know where the growth is going to come from tomorrow. And so when I look at something like ads, that was something we were building towards. Consumer hardware is something we can think about next year. And so there's a lot of reasons to think about OpenAI as a long-term, durable, strong business.
EISEN: How are you thinking about when to go public?
FRIAR: Raise money when you can.
EISEN: Which is what you're doing.
FRIAR: What I do, for me, one of the underpinnings of going public, beyond just raising capital, which we clearly have been able to do at scale in the private markets, it also comes back to our mission of AGI for the benefit of humanity. And I personally feel very driven to put that economic value creation that's happening into as many hands as we can. Like, why should it just be limited to institutions? And so I do get excited about public offering from that perspective. But that's in our future. And today's just about managing the business.
EISEN: 2027 story?
FRIAR: I think last we said we would continue to look in '27.
EISEN: Does the midterms alter anything about this plan?
FRIAR: No, I think, from a midterm perspective, look, we are, we're paid as operators to manage scenarios. I don't think there was something that dramatically changed for us. Again, it comes back to what I said. If the midterm means more focus on data centers and communities, we're going to keep doing the work to get into those communities and to build trust with them.
EISEN: And you're raising money, but is the capital there? I mean, what's the biggest constraint? There was a report yesterday that $500 million in loans got, got pulled, basically got sold, including a loan for the New Mexico build-out of Stargate. So, is it a capital constraint? Is it power constraint? What is that, what is the big headwind?
FRIAR: I think capital will always be available for the strongest stories and the most exciting stories. And this is kind of where we started. The demand for AI is real. I don't think anyone's arguing with that. Or, if they are, I think it's a losing argument. We see it every single day in the build of the way people are using our products. And, to me, those are businesses that investors will always want to invest in.
EISEN: OK, well, Sarah Friar, thank you. It was a, it was a pleasure going deep into everything we have learned this week and everything that's going on with you in the business. So thank you very much for taking the time.
FRIAR: Thank you.
EISEN: Sarah Friar, the CFO of OpenAI, with some real answers, Carl, and new information, I think, on how the business is doing and how she's thinking about raising money and the path forward.
For more information contact:
Stephanie Hirlemann
CNBC
