How Loews Hotels Puts AI to Work for Guests and Teams - Dan Kornick, Loews Hotels

Josiah: I am curious, Dan, because I was looking into your background. You've worked at a lot of interesting places, done a lot of interesting things. One of the things on this show, I'm trying to give people a sense of why work in hospitality, of all the industries, of all the things that you could do. And I'm curious for you personally. You have a really interesting job, but why apply your talents and your energies into the hotel business as opposed to some other industry?

Dan: That's a great question, actually. It is a people industry at the end of the day, right? It is serving people, putting them on vacation. Where do you get a job where you get to really send people on vacation, make sure they have a great experience, a great time? How do you impact that? So to me, that's the part I love about hospitality and my job in technology, right? 'Cause it's not transactional. It's not a banking, it's not a healthcare type of job. But how do you make people's lives more enjoyable? And that's the part I enjoy the most.

Josiah: I love it. It's a great answer, and it's so funny on the show. I'm talking to people in different parts of the hotel business, but it all comes back to that, right? I feel like wherever you find yourself working, it comes back to that. It's interesting to think about that through the lens of technology. So we're gonna have a great conversation today about technology, but it's interesting that is your North Star, that's your motivation.

So I'd love to jump into it if we could. I really enjoyed your remarks at the 2026 NYU IHIF event. You were on stage talking about some things that you've done. You talked about the importance of people, the importance of how this technology supports all of this. It was my favorite session there, and I took a bunch of notes from that session. One of the things that stood out that might be an entree into our conversation was this notion of starting with data. If my notes are correct, I think you said something to the effect of AI is only as good as the quality, the accessibility, and the connectedness of your data and the systems underneath. And of course, there's a lot of applications of AI. I wonder if we kind of begin the conversation there. Why start with data? Tell me a little bit more about that.

Dan: I think you just summed it up already. It is the foundation of AI to begin with, right? And I think at Loews we're very fortunate. All of our properties use the same systems and technology. We're managing them all centrally, so this allows us to do things uniformly across the brand, and data is one of those, right?

Now, for the past few years, we've been focused on kind of a simplification approach, consolidating our systems, consolidating our data, consolidating our interfaces where it makes sense. So this has allowed us to eliminate that duplicate data. It allowed us to eliminate complex interfaces and make things simpler. So as an example, this year we've consolidated our PMS, CRS, and guest profile data onto OPERA Cloud Central, so I now really have one central place to operate from. And I think when you look at that, it just makes it much easier now to apply AI on top of that, because you're not dealing with conflicting data, you're not dealing with conflicting systems or things that are out of sync.

Josiah: So Dan, I gotta jump in here and ask you about this, 'cause I was just talking the other day with Professor Chris Anderson at Cornell. I always find him an interesting voice, sort of impartial in terms of what's going on. And one of his points was he anticipates there's gonna be more consolidation away from point solutions into platforms, and what you just described seemed to support his theory. He expects more of this in an AI world. Talk to me about that decision to centralize, moving in this direction of centralized technology.

Dan: I think it's simpler, faster to roll out change across all of your different platforms. I think your data is consolidated. I think your business processes are consolidated. So the pendulum swings from point solutions to really those platform solutions. Every probably five or 10 years, that becomes a debate. I just see it, given the cost standpoint, the complexity standpoint, and being able to really utilize the data and the systems better. We're at that point now where the pendulum's swinging towards consolidation and platforms versus having hundreds of point solutions.

Josiah: Interesting. So I'd love to dig into this notion of data a bit, and that idea of the quality of the data being important. There's a couple different components you hit on there. When it comes to data quality in hotels, I imagine the systems and the way they catalog that is a piece of it. But what are some of the complexities that you had to overcome when it comes to data quality in the hotel business?

Dan: I think the complexity is when you're having data, whether it's customer data across multiple systems, trying to keep them in sync. Whether it is information about your properties across multiple systems, trying to keep them all in sync, trying to keep them all updated consistently. There's a lot of manual processes that go on. There's a lot of auditing that can occur, right? So I think the more consolidated, the more clarity that you have on data ownership, data governance, and really a single ownership and focus of data is key.

Josiah: Interesting. So there's a ton of data in the hotel business. I think what you are creating is gonna create a lot more data. When you think about your data strategy overall, I'd love to hear your thoughts on this. Where does all this data live? How do you think about maintaining its quality, its usability, its accessibility, that sort of stuff?

Dan: Could probably go on for about an hour on that question alone, right?

Josiah: Let's dive deep. I love this stuff.

Dan: Well, as I said, our fundamental tenet was let's get the foundation right. Let's get all of our data consolidated as much as possible. Let's make sure that we have clear ownership, that we understand the processing of that data from beginning to end. So I think that has really helped in our data strategy from an operational side, right? You're never gonna be perfect, 'cause you always still need multiple systems and to bring it together, but that's key. For the reporting side, we've brought a lot of it together also to enable our reporting. But now with AI, as long as you know where you're pointing your different data to, it doesn't have to be centralized. You just have to be clear on where you're getting that data and how it's being used.

One lesson that we learned in this was recent. We have a Chat Your Service program where we can answer guest questions or respond to any requests you need, whether you need additional towels, you would like to order room service, or you wanna know where the best ice cream is in New York. You can reach us seven by 24 via text. So very easy to use for our guests and very efficient.

But what we noticed, we had rolled out AI as a pilot at one of our three-star properties in Orlando, and we had pointed AI saying, "Look at our website for information. Here's where we want you to look for information." But people were asking about restaurants, and the AI was now looking at reviews on other platforms outside of our data, and it was actually giving information about a restaurant we had closed down many years ago.

So what you have to understand is you may control your data, but there's also data out there that AI's gonna grab, and how does that come together? So our solve for that is we've gone on our website and put in a hidden section things we also want our AI to read and understand. "We closed this restaurant three years ago. Please don't prompt that up as an option," as an example, right? 'Cause AI's gonna look at your data, but it's also gonna look at everything that's out there on the internet.

Josiah: It brings up for me this idea of guardrails, and how do you make sure that you're delivering on what you're intending to deliver? How do you think about parameters or guardrails or harnesses for AI to ensure that it's delivering on what you need it to do?

Dan: I think you need to be very specific on where you're focusing it. And as I just said in that example, you're never gonna get it right. So you always have a human in the loop. You're looking at what's occurring, you're understanding the results. You're understanding then how do you put those guardrails in place as you learn.

Our approach is very pragmatic. We wanna start small, learn from that, right? 'Cause there's a lot of learnings as we go through this. And then we'll slowly increase or broaden those guardrails the more we learn, the more that we're confident that what we've done is gonna give the right answers.

Josiah: So one of the things that stood out to me from your remarks on stage at NYU back in June was you were talking about this notion of escaping pilot purgatory. It sounded to me like there were a lot of experiments happening as you've been on this AI journey, testing the use of AI and what benefits it provides the organization. I would love to hear about how you selected winners and doubled down. But taking a step back from that, how did you think about pilots? How did you think about testing things across the business in the early days of AI at Loews?

Dan: Let me answer that in a different way. Let me talk about our approach to AI. We look at it in kind of a progressive manner. So our approach to AI is focused on, first of all, leveraging our current vendors' capability, whether it's to solve material business issues, drive a better guest experience, or more efficient operations.

So listen, we're a hotel company at the end of the day. We're not a technology company. I don't have a large development team. So I wanna be able to leverage practically what my vendors are putting in place that sits on top of our data that's already in our current environment. It's easier to govern, it's easier to deploy, and it's easier to get a practical benefit from.

Now, if that isn't present, if I can't leverage that with my current vendors, then I'll look at other vendor solutions that would have an AI-native capability. And really, our third solution, if I can't solve that problem, we'll look at AI agents once we've exhausted those. But so far, we haven't had to go to developing our own AI agents. We haven't found a big enough business case where that's made sense so far.

'Cause our philosophy's pretty simple. We're going to be fast followers. We're a hospitality company at the end of the day. We need to be competitive with our peers, both at a top line and a cost. We need to make sure AI drives value, right? Technology needs to go beyond neat or cool. It needs to drive value that you can see results on a balanced scorecard, not really be an impediment or a detractor from what we're trying to achieve overall as a business, right? And we wanna make sure we're learning from it, that it's giving the right results in a repeatable fashion.

So our focus isn't really doing a ton of pilots necessarily. It's been trying to layer on what our vendors are developing in the space, which are key partnerships for us, and how do we use that to drive value.

Josiah: So that's fascinating, Dan. I mentioned off air I've been working with Naz at Destination AI. We'll both be there. I encourage people who are attending to make sure they attend your session. But some of the research has been fascinating. It supports what you're describing, because there's a mix, it feels, of approaches to using AI. Some are working directly with the frontier labs and kind of building their own solutions. Others work with their technology partners first.

And what I heard from you is this framework of start with your existing technology partners and their capabilities. If they can't do it, other hotel technology providers, and then you look at the custom builds. What was interesting in the research I've been running is the ones that are using the approach that you're describing are reporting better payback, better ROI, as opposed to the ones that are doing custom builds. So it's interesting to see already. I know we're early days, but this approach seems to be a smart one.

And when you think about working with your vendors' capabilities, you mentioned that it's easier to govern, easier to deploy. I think that connects to our earlier discussion about data, right? This all comes back to data. Where is the data ready to be used in a trusted, safe way? And it's probably with your technology partners that you've selected, right? So it's interesting to hear that approach.

Dan: Exactly. And I think it's more of a guaranteed way. By the way, I get 100 calls a week from different AI vendors, different technology vendors. Everybody can solve every single one of my problems with AI. I probably spend more time just looking at AI solutions, but yet we're not gonna pilot a lot of things, 'cause we wanna follow our philosophy. What really is gonna make an impact? 'Cause you could lose sight of what you're trying to do as a business by just trying to experiment with AI, and that's what we're trying to avoid. So if we leverage what we already have in place, as you just said, the governance, the guardrails, the data, the ability to execute quickly is all there.

Josiah: It's so interesting. I love what you said there about we can't lose sight of what we're trying to do as a business, and I think you've touched on this a few times in our conversation already. As you think about the capabilities that new technology provides, as you think about your North Star, you're here to serve guests, you're here to support your teams. Tell me a little bit more about that. What are you trying to do as a business, specifically as it relates to technology?

Dan: So if I look at what we're doing today from an AI standpoint, I'll follow along with our practical approach to that. I'll start with team members, and I'll walk through each facet, right?

So from a team member side, we allow all of our team members to use Copilot Chat, to use Claude, ChatGPT, any tool they want to help them learn, understand, and drive individual productivity. Now, for us, Copilot Chat's the only one in which they can use business-sensitive data. They can't with the other tools, but we allow them to use the other tools.

We've done a pilot with about 70 people that had the full Copilot license, and what we found out is after a three-month period, it's really freed up maybe six, seven hours a week, where they can now allocate more strategic time. We don't wanna roll it out to all of our thousands of employees. It just didn't make sense for the cost versus the value, right? Once again, pragmatic approach. We rolled it out to 70 people. We measured the results. We'll put it out in cases where it makes sense for people to get that additional value. So that's kind of the team member, individual side of AI.

Josiah: So hours a week are being saved, and I was curious how you measure the impact of this. Is it you kind of check in with them, like, "How many hours did this save you?"

Dan: Right. So we went through a 12-week pilot period with about 60 team members. We asked them to record where it saved them, specifically time. Where did it save them time? Where did it save them effort in their current jobs? Some were very small, some had some great use cases in which we were able to save time. So we measured that to see what is the true benefit of this, right? 'Cause there is a cost to it, and we're cognizant of the cost with AI, because I've seen a lot of solutions where the cost actually outweighs the benefit. So we're trying to be very pragmatic on that approach.

Josiah: So that's fascinating. You're testing, you're measuring, you're being pragmatic. And you mentioned there's a lot of stakeholders you're thinking about. So there's the team member benefits that you're thinking about. Tell me about the other parts of the approach here.

Dan: All right. From a guest services side. So currently, as I talked about, we have Chat Your Service, where you can text us 24/7 at all of our properties. Right now we have people answering that. But we've put AI in place now that recommends replies to our team members, so it's kind of their personal assistant. Those replies are gonna be more personalized, 'cause AI has purview to all that data about the guest. It's gonna be more consistent, really a faster guest response. So that's where we're kind of putting that AI next to our team members to really help them do that.

And once again, we've piloted full AI replies for one hotel in Orlando, one of our three-star hotels, so that we can learn from that. That has gone very well. We've gotten a lot of good guest feedback in terms of responsiveness, correct answers, and such. Once again, we're still in the loop. We're looking at how AI responds and really training the tool better.

On our website, like most other companies, we have a web chat with Loews, so it answers probably about 75% of the questions people have about our properties, amenities, restaurants, things to do, things of that nature, right? But what we put there also is an AI agent that gives you the ability to retrieve a copy of your folio. A lot of us who are business travelers obviously need that folio for our expense reports. We get thousands of calls a month into our call center where people are requesting a copy of their folio. They lost it, don't have it anymore, right?

So we've been able to automate that, where AI will ask you one or two questions, validate who you are, and you have your folio in about two seconds via email. So to me, that is a great opportunity where AI is providing an ROI, but it's also providing a better guest service on something that's more of a lower... value is not the right word, but a very transactional element, right? It's not taking away the human touch. It's not taking away the value that Loews provides. It's just giving you a faster way to get what you need to get done.

Now, coming up for our call center, we're also looking at some voice agents. So we're gonna be piloting those next quarter. Obviously, we're not gonna start taking all of our calls with AI voice agents. That's not who Loews is. But as an example, can we handle transactions where you're just canceling your reservation? Can we do that with an AI voice agent where, once again, it's more efficient, it's a very simple process, and it's something that's more transactional? So we're looking at where AI can come into those places that are better for the guest and better for us, right? So that's from a guest side.

From digital marketing, we've implemented a new customer data platform with AI that's part of that platform. So now our digital marketing teams are creating thousands of targeted marketing campaigns to our customers. That's something we couldn't have done before at scale. We're doing it so much faster. Where it would have taken weeks before, we're doing it within hours, right? Now, this obviously drives incremental revenue, but it's a better marketing experience for our guests. It's more targeted. It understands who they are better. So that's another example.

For us, we build properties. So we are an owner-operator at Loews. We build our own new properties. We've put in an AI-powered construction management tool. So this allows us to really look at all the scheduling of all the subcontractors in conjunction with our general contractor. We've been able to optimize two months of a construction build using AI.

Josiah: Wow.

Dan: Now, listen, when you're spending hundreds of millions of dollars on one property, there's a big ROI in getting that done on time and on budget. So I think that's been a nice place for us.

Josiah: It really stands out to me how pervasively across the business you are deploying AI and getting benefits from it, right? I think construction is one I don't hear often in the AI and hotels conversation. I'm curious for you, Dan, though, overseeing all this, is your life any better with AI? You get 100 calls a day, everybody's trying to talk to you. I imagine there's so much to do across the business to support your teams. Has AI made your life any better?

Dan: Has it made it better? It's made it different. It's made it more interesting. I can't say it's made my life better yet, right? Because I'm doing my day job, and I feel like I'm spending another 50% just focusing on opportunities, potential.

Josiah: Yeah.

Dan: Making sure that we do the right things, that we're thinking ahead, that we're thinking strategically, that we're thinking about the security, the governance, the rollout, the learnings. How do we all apply that in something that's still fairly new to all of us, right? But that's what makes it feel exciting. That's what makes it fun. I can't say it's helping me necessarily, but it's keeping things interesting, we'll say that.

Josiah: Ooh, I love it. I mean, if it's more interesting, if it's more engaging, that's also... You're definitely not alone. I think this is coming out in the research, too. There's high excitement. The world is changing quickly. For leaders overseeing all this, they're not necessarily saving time, 'cause there's almost more than ever to do. It's awesome that your team is starting to save time. That's great. But for leaders, that is very common, this mix of all the things. But I think more excitement is also a good thing, because you've seen technology through different innovation waves across businesses. So if you're excited about it, I feel like that's also a good outcome as well.

Dan: Oh, very much so. Yeah. I think it keeps our teams engaged, our business more engaged. And I think our biggest obstacle, though, is just expectations and perceptions, because everybody's reading every day about what everybody else is doing, and everybody has this false perception that they're behind, or that they're not doing the right things, that they're not thinking enough.

And I think in the hospitality industry, we beat ourselves up. We think we're always behind. We're not on the leading edge. Well, we're never gonna be on the leading edge with technology. That's not who we are in hospitality. But I do think as an industry, we are applying AI in a pragmatic way. I think we're applying it in a way that makes sense from an owner, an operator, and a brand standpoint, right? You're seeing the benefits at all three levels, and that's sometimes hard to do in our industry, by the way.

Josiah: Very.

Dan: So I think you're seeing that benefit in that, but I still think the biggest obstacle is perception and expectations. Once again, I get hundreds of outreaches a day. Everybody can solve every single problem that I have, and that's great. If they all worked, I wouldn't have a job, which probably wouldn't be a bad thing. So that's just one of the interesting things that are occurring right now with AI.

Josiah: But I think it's a fresh take, 'cause I've heard a bunch of people come on the show and talk to me about that dynamic, and they'll bring up some consulting chart showing the adoption of tech in the industry. And so it's interesting. I think this is what we touched on at the beginning of the conversation. You have a North Star. What are we trying to do here? What does good look like? I was gonna ask you how you cut through all the noise and stay focused. And it seems from the conversation that we've had so far, part of it is fundamentally, what are we trying to do as a business? What is the role we believe technology can play in that? Is that useful in cutting through the noise?

Dan: That is. And by the way, our entire senior leadership team has that same philosophy. So I'm not getting a lot of, "Dan, you need to look at all these things." I'm getting the, "All right, what is pragmatic? What is practical? What are we trying to solve in terms of business problems?" Not, "What's cool or neat with AI?"

Josiah: Interesting. So you said the whole leadership team. When you're having a conversation with your CFO, walk me through this. You mentioned costs can get out of control with AI, and this is something I feel like is not talked about enough. Just throwing AI at it actually can cause costs to spike. So that's a dynamic. When you're talking to your CFO, what are those conversations like? How do you evaluate investments and things like that?

Dan: Oh, he is very cognizant of the cost, and he's reading a lot of articles where, hey, you see token usage getting out of control. You see vendors now changing their business models when they implement AI, and how you get charged for that AI usage. So every opportunity we look at, we also look at what is the cost, or what are the cost opportunities to really get out of control? 'Cause once again, it'd be easy to just implement AI in 100 places. But if I'm now getting hit with costs that don't really have a clear business driver, if I'm getting hit with costs that I really can't contain or control easily, then we're not gonna do it. Doesn't make sense, because the value's not there from the business side. We focus on that North Star.

So we are looking at, okay, so from an AI standpoint, how am I gonna get charged? What is my volume? How is this gonna work? And each vendor is changing their model almost on a monthly basis. I can't say specific vendors, but I've looked at, hey, this is a great opportunity to implement AI, but for what they wanna charge, I could just hire three people to do that. That's more cost-effective, right? Or, hey, you can do it, but now you're gonna get charged these tokens, and you really can't control it. All right, so that's not something we're gonna jump on right away.

Josiah: So this is interesting. It's also timely, Dan, 'cause I think a lot of leaders are going through this budget process, and they're thinking about how do I plan, and it feels like the world is changing every week. And it feels like, to me, thinking back to your framework, working first with your technology partners, that is a guide. You mentioned predictability. What does that look like? If I heard you right, it also sounds like some technology companies are rethinking their pricing all the time, which makes it very difficult to plan. So it sounds like you kind of need a predictable model. This is the cost. It needs to connect to a business outcome. SaaS, I feel like, was okay at that, and token spending and usage is much harder to calibrate against. Is that sort of what you're seeing out there?

Dan: That's exactly what we're seeing out there.

Josiah: Yeah. Interesting. So what I'd love to get your thoughts on also is all the ways of accomplishing things with AI. You mentioned token use and not only different economic models, but also driving business performance. I was listening to Microsoft CEO Satya Nadella this morning give a talk about what he sees from his perspective, which is interesting 'cause he's not running one of the frontier labs, and he works with a lot of large corporate businesses. And it seemed to him that it's really important to make sure that you own your business processes. You have kind of sovereignty over that. So if a frontier model pulls the rug on something or changes the business model, you're not high and dry.

From what you've shared, it seems like that's kind of what you're thinking about. You're building these processes, the data that you own. So it seems like it's giving some resilience for whatever the future might hold. Am I hearing you right there?

Dan: No, I think you are hearing us right. And when you think about it, AI's not gonna fundamentally change the hotel business model, right? Even if I look at it today at Loews, we've got over 11,000 team members, but most of them are frontline workers focused on housekeeping, guest services, food and beverage, banquet servers, lifeguards, whatever it may be. AI's not gonna be a big factor in their jobs or in how we deliver to our guests, right? Because it's still a people business. It's still an operational business.

If you were a tech-only business, I think you're looking at probably different risks, right? But from a hospitality side, I don't see it being as dynamic from a risk standpoint, 'cause once again, I'm not gonna use AI and all of a sudden replace our lifeguards with AI, right? If you're drowning, I don't think you want an AI to come out and save you, right?

Josiah: Yeah.

Dan: Or I don't think you want an AI to be serving you a great steak dinner. So I think our industry's a little different, probably a little more resilient in that standpoint because of how we operate and what we do.

Josiah: Interesting. So one thing I'd love to get your take on as well is, you brought up this idea of human in the loop before. I know at NYU you talked about this a little bit, and it's interesting with this Chat Your Service or some of the pilots that you have coming up. Tell me a little bit more about this. I've heard about this concept. What does it look like practically to have a human in the loop, maybe in a guest-facing, guest service sort of AI environment?

Dan: Well, what we're trying to do is make sure that AI's there to support the team member doing their job, right? So if I'm in the call center and I'm recommending what you could respond to, or I'm learning from what you do respond to, to help you respond better in the future, right? So to me, that's more AI is in the loop. Where we now have AI fully answering things, as in a pilot stage, we're now reviewing its output, right?

And we're not using AI where AI's ever gonna make any key business decisions for us. Doesn't make sense in our industry and what we do, right? So we'll never use AI in terms of HR decisions, key business decisions, things of that nature. I think we're just being cognizant of how we use it, what the risk is in using it for each case, and making sure that a human is overseeing what is occurring, either AI supporting that human or a human overseeing what that AI is producing.

Josiah: I feel like we covered so much in our conversation. And I'm curious, we have the Destination AI conference coming up, really excited to see you there. But what's on your mind? I feel like there's so much going on in the world of AI. You're building incredible stuff. What's on your mind? What else is exciting for you these days?

Dan: Great question. What's on my mind in AI specifically, I'll start with that, but in general. Just in the past week it's been crazy. You've had three CEOs of three different AI companies come out and state that we need to slow down, we need to have guardrails, we need to do it in a more safe environment, right? You had the Chinese government pretty much echo the same thing over the weekend. You've had OpenAI just even yesterday come out with, "Hey, we've seen some issues with our newest model," right? You've seen all of the security issues with Mythos coming out.

So I'm seeing all the benefits, we're seeing how we use it from the business side, but it's always still, what don't I know? What is truly going on when you have all these people urging caution? What are they seeing that I'm not? What don't I understand? Where's it going? So those are kind of the things that are on my mind. It may not be the fun things, but it's just trying to understand what that is and how that's gonna impact us as a society. What are we doing as a business? How's it gonna impact what my kids are gonna be doing for a living in five years?

Josiah: We're all thinking about that. But I think, Dan, what's interesting to me is it feels like it's important to think about where this might go. What I've also heard from you is you're acting on what you do know. So you do know that we need to take care of guests, we need to support team members, we need to create an AI-enabled technology environment where you have some control regardless of where things go. It feels like you're acting on what you do know, building on the best of technology today, and I feel like that's cool to see. Even amidst uncertainty, you're not saying, "Let's not do anything until we figure this all out." You're moving forward with what we do know, with cautiousness.

Dan: Right. 'Cause if you waited to figure it out, right? But we're also trying not to get over our skis. We could easily be piloting a lot of things. We could easily just be spending a lot of money on things that really aren't going to move the needle for our business.

Josiah: Love it. Love it. I feel like we covered a lot here, Dan. Was there anything that we did not talk about that you were hoping we would talk about in this recording?

Dan: No, I think those are the key things. So it's been a fun time talking with you today.

Josiah: Likewise. Thank you.