Digital Skies: How Technology is Reshaping Latin America's Aviation Landscape

Transcript

Charles McKee:So we touched on it yesterday, but we wanted to go really deeply into the discussion today to unpack really what's happening in the industry around artificial intelligence. So I'm very pleased to be joined today with this august group of panelists. To my left is Marc Lumpé. Chief Operating Officer of Air Transat, Canada's leading airline and travel company. Next to Marc is Maximiliano Vicini, who is the Head of Business Development at HITIT, one of the world's leading passenger service solution providers. Next to Max is Viktoriya, Viktoriya Soubra, who is the Director of Sales at iCoupon. And then Last but not least, we've got Eric Keeley, who is the Senior Director of Product Development for Flightscape at CAE. So we've got a variety of different viewpoints. We've got a lot to cover today, so let's sit down and unpack it. This is really a discussion that involves you as well, so we'd love to hear your voices either through Slido, of course, or at the end we'll leave a couple minutes for some questions and hopefully some dialogue. This is a topic that affects all of us. In fact, actually, I'm reminded yesterday it was Manuel van Oort who had a very interesting little perspective. He said, AI is fit for any process. Every single process is absolutely optimisable. Which, do we believe with that? Do we agree with that? Let's unpack it and see. Eric, I want to start with you. We were talking yesterday about other significant transformative changes that have affected the airline industry and technology. You had a couple observations. If you were to think about before the coming of artificial intelligence, what has been one of the most significant things that have really rocked the industry?

Eric Keeley:Well, first of all, thank you for allowing me the privilege to be up here to talk about technology. I'm very excited about it. Yeah, we did have that discussion. Let me ask in the room, how many— do we have any licensed pilots or has been a pilot before? Raise your hand.

Charles McKee:Any pilots?

Eric Keeley:We've got a few pilots. So you recall having to, at one point in time, haul around a giant bag full of navigation charts to your aircraft, usually between 50 and 70 pounds. I don't know what that is in kilograms, being from the US and all, but these— there was never a time— it would have been unheard of way back then, 15, 16, 17 years ago, that we would allow an iPad to to replace all of the navigation that—

Viktoriya Soubra:An iPad.

Eric Keeley:That's right, an iPad. Now there's iPads throughout the entire industry. Every aircraft is basically utilizing iPads to replace critical navigation data that's guiding the aircraft. That, I think, was— we never thought that that would be possible. It does prove that our industry can change and can change rapidly when were motivated to do so. I think that was an incredible advancement in technology for us.

Charles McKee:The iPad was a big change. If you were to think now about what we're seeing unfold with artificial intelligence, potentially affecting every process that an airline in every domain may need to rely on, exponentially, how big is this? How big is—

Eric Keeley:I think you multiply this by 100 or maybe more. This is actually going to be incredibly transformative across every part of the airline, including from the flight deck all the way down to how you get your tickets. I mean, this is going to be an incredible transformation.

Charles McKee:Let's talk about this. Let's talk about some of the use cases. I think, Marc, I want to turn a question to you, and that is, in your role in operations, you have a really good 360-degree view of what's happening in the airline. When we think about artificial intelligence, not so long ago, typically conversations would start with, well, we're doing a chatbot for our contact center. Here we are, not too far beyond that today. And when you think about the kind of interesting use cases that your company is actually exploring, or use cases that interest you, what are those? What are some—

Viktoriya Soubra:Yeah.

Marc-Philippe Lumpé:So I think this, what you mentioned in the beginning, I think it's true. In the next couple of years, the whole structure of airlines is going to be completely changed upside down. And as currently we are still seeing that you're screwing AI and other solutions on top of existing processes, this is going to change radically in the next couple of years because you you will have to think about this differently. It's going to be AI at the core, and everything is going to be around that core, which is going to dramatically change the way we're going to work, because nothing is going to remain as is. Nothing. For legacy airlines such as Air Transat, this is a big challenge, because if you're a startup in a couple of years from now, you're just going to implement your business processes, Right away with that mindset. We have to maintain an operation today while we're doing this transformation. And that means that you have to choose your— progressively choose the departments that are ripe for these changes. Very far advanced on the operations side is already maintenance. We are, as we speak, nearing the end of a 3-year transformation transformation phase in maintenance in Air Transat. We are going to be the first airline in Canada that is going to be 100% digital. There's no more paper, nowhere in the value creation chain in Air Transat as of the end of this year. And with this, we are implementing an AI solution for predictive maintenance because now everything is accessible to that. And the maintenance department in Air Transat is going to be the first department which is going to be ripe to flip into an AI-centric mode. Others, it's going to be probably the integrated operations center, and then you move your way progressively forward because you have to see how solutions become available. Also on the commercial side, there are a lot of very interesting solutions surfacing now. But there the integration tends to be a little bit more difficult because software solution like PROS. They don't have a real interface for these AI solutions. So what you will see over the next years is as the technology evolves, you will pick different departments, different divisions, and then you're going to implement true AI solutions. Because this chatbot stuff, it's all nice and beautiful, you have to do it, but that's already to some extent a thing of the past.

Charles McKee:So a couple things, a couple takeaways there, and these are— hopefully we'll find throughout this discussion certain Practical lessons, lessons that we're experiencing. But one of them, I think that what you've said was accept the fact that AI will be at the center of virtually any process, any business process. But what's also important is I think your experience in technical operations is there's almost a prerequisite. You're digitizing the operation and then you're really infusing it with AI.

Marc-Philippe Lumpé:Yeah, you have to do it that way around. So you have to get your basis right first. Which was essentially the first two, two and a half years of this of this transformation because digitalisation, digitalisation is the base of everything. So you have to get everything into the systems to make it available to AI. And if you try to to jump that step, you're not doing yourself any favour because then you're going to be half. Then you will have a half cooked solution and you will never be able to. really leverage all the potential that that has. One step at a time. It's very important to be thorough when you do that.

Charles McKee:Eric, you look like you wanted to say something.

Eric Keeley:I think that's very valid. Marc is— when we've talked before, he's very visionary in terms of how he wants to transform. What I find is the difficulty that we have is trust and trusting in the technology to make that kind of a leap. from, say, the rest of the airlines in the industry. So that can be a very difficult leap to make, is I'm going to completely transform and digitize everything and trust that AI can solve these problems.

Charles McKee:Let's hold that thought, because I think that when we talk about deployment, that probably falls into the category of what are some of the challenges that we, either from a mindset perspective or a technical perspective, need to overcome. I think one of the things on the technical front that I also want to touch on that you mentioned is the idea of having an API, to have an AI-native API available to be able to do what you want to do. You've mentioned a vendor that needs to have an API available to you, and we'll talk about that as well. I think these are really important enablers These are topics that are enablement for deployment. One of the things that touches all of us, I think, in this business is, of course, what happens when there's an irregular operation. It costs airlines lots of money. It's a lot of loss of goodwill sometimes with customers. There's recovery processes, et cetera. Viktoriya, we were talking a little bit about this topic. I'd like to ask you to just share a little bit. Your experience at iCoupon, and you guys are really at the heart of, of Europe recovery. With your platform, I understand that you've done some very good work in enabling the vouchering platform with fraud detection, compliance with some of the regulatory requirements, using AI to do that. Can you speak a little bit about how has your experience to date both impacted the business performance but also customer performance?

Viktoriya Soubra:Just for those that may not be fully aware, iCoupon is a digital vouchering platform. And what we do today for airlines— and about 190 airlines today are currently using iCoupon— is to digitize their meal processes during the disruption. In this region, we have Resolution 400 that is very strict when it comes to having to provide compensation In Europe, we have EU 261. Other countries will have other legislations that require the airline to provide meal to the passenger. When we're talking about issuing vouchers via paper or virtual credit card, it is impossible to comply with the regulation. What iCoupon does is we utilize a sequence number as a unique identifier. What passenger already holds in their hand when they're in the airport being delayed is their boarding pass. Issuing an entitlement onto a boarding pass during the delay ensures that airline is 100% compliant when it comes to irregular operation and legislation requirements. It also means that 100% of the passengers will receive that voucher. It's just a matter matter of communicating to the passenger that they have an entitlement. What we've seen, especially in this region, which is very interesting, LATAM has been our very first customer and they started digitizing the processes with us in this aspect 6 years ago. It is only starting from probably mid-2024 and 2025 that we've seen majority of the airlines locally in Latin— in South America that have moved forward and digitized with iCoupon. They have recognized, well, they came to the party a little bit late. As I said, we're already working with 190 airlines. On the upside, they had the opportunity to see who has the global technology that is being utilized by everyone so they can adopt one of the best ones, the ones that actually work to deliver what needs to be delivered to the passenger. In terms of the compliance, we provide that to the airline. In terms of the convenience for the passenger, again, they have nothing to download. They don't have anywhere to go. They just need to be informed about the fact that the voucher is— or their boarding pass has become their voucher. Wherever they are in the airport, they have more time to spend. They have choices. They're no longer limited to a sandwich and a water as such. They can go and purchase whatever it is that they like without having to cause additional disruption and queues at the airport. While we are supporting the airlines and the passengers, we are also supporting the airports, especially during mass disruptions. Thousands of hours of admin is removed from the airline operations with our digital solution, as well as airport order is restored in such a way.

Charles McKee:Artificial intelligence. What would be some of the practical benefits that a customer might expect? As you think about improving your processes, is it faster delivery, more efficient delivery? What are some of the— What are some of the tangible benefits that you think that artificial intelligence may provide your customers?

Viktoriya Soubra:I believe the artificial intelligence will speed up the process of delivery, communication to the passengers, and the ones that are more tech-savvy and frequent flyers, they are accustomed, they know the rules, they know the regulation, and it's much easier for them to manage their journey in the airport during the disruption especially. However, I believe that there's still going to be 10%, 15% of the passenger that will require personal support. However, removing majority of the work and admin with AI for majority of that flight and only dedicating the work and attention of the airport staff to those who actually need it, that will just mean the loyalty of the passenger to the airline, convenience, and happy experience.

Charles McKee:That's a very interesting point. The idea of actually being able to segment the type of service you give based on what the actual need of the customer— more responsive, it's more personalized level of service then. Let's go up a step though, and let's talk about the root cause of why iCoupon is so successful. And that is, there are irregular operations out there. When we take some of the comments, Marc, that you speak about in terms of the use of AI for predictive maintenance and other things you might be doing in the TechOps area, do we see this as being a significant driver in reduction of AEROps for the future?

Marc-Philippe Lumpé:Well, yes, obviously you will see that because in— well, first of all, you can already avoid some of the delays in the first place because in predictive maintenance you don't have then the disruption per se. But let's assume you have the disruption, then it's always about generating solutions. And the generation of solutions is that complex that it is just inaccessible to the human brain to find an optimal solution. And there is AI really coming into play. The problem that we have faced in the past was twofold. Number one is you have to optimize aircraft, crews, and passenger solutions in parallel, which is a complexity in itself. There have been historically some systems in the market that claimed they were able to do that. But beyond that, the problem is that in IROPS you sometimes have to break certain rules. So you have to change reality, which means, for example, if for a solution to be implemented I have to go below minimum connecting time, then I have to potentially make that happen. And that means that the systems have to be able to reasonably break rules when they propose solutions. And this may sound Simple.

Viktoriya Soubra:Yeah.

Marc-Philippe Lumpé:But from an IT perspective, in the past it wasn't possible because for a computer, 1 1 is always 2 and never 2.5. And that doesn't work in IROPS. So you need to have an interaction between the, the AI agents, because there are multiple that are negotiating and developing solutions, and then at a later point of time with the human agents in the operations center. There has to be an interaction because they have to make the solution happen by changing actual reality. And this is going to be, in my opinion, at least for Air Transat, as far as I'm seeing that, that's going to be the second department that is going to be AI-centric because everything is available digitally. Number one, important. And number two is, I think we are— I believe we're having now a potential solution that is going to be able to to do what I was just describing. This is what I've been saying since, I think, about 10 years. People who know me know that I've been going over this again and again. This is going to be the holy grail in IROPS if you can do that. This is going to drastically change the quality and cost of IROPS resolution.

Charles McKee:This is good. This is good. But now I'm going to go back to Eric. Eric, you talked about trust. Trust in the recommendations, trust in the the output of artificial intelligence-enabled processes. Based on what Marc said, if you had the easy button— does everyone know what an easy button is? Easy button is a red button that you can hit and it just happens. If you had the easy button in front of you, what are some of the considerations in terms of trusting the output for you to hit that easy button?

Eric Keeley:Yeah, let's talk about that. That's a really great question. But I want to put in context what does iRoute mean to the airline industry first, and that is 30,000 flights, over 30,000 flights every single day are disrupted, are delayed. That's 1 in every 5. That is a huge problem across our entire industry. Between 5 and 8% of every airline's gross revenue is being lost to disruption. And when we did studies on this, 50% of those, we believe, can be impacted in a positive way. In other words, reduce the impact of the disruption or avoid them altogether. And so, of course, if I had an easy button, why wouldn't I hit it? And so when I go to customers and I ask them, hey, would you allow me to put an AI agent in that will solve this problem? Invariably, the answer is no. And I know that may be not intuitive to you. Of course, why wouldn't I do that? Because the AI is doing things that is not understandable necessarily to the operator who's responsible for those flights, who's responsible to make sure that that flight is going to get back on track. And so if they just hit this button and it does all sorts of weird things and it doesn't understand it, the trust is not there. They simply are not adopting them. In my experience, we have had solvers in the industry for a decade, and they're not using them effectively. I think we have to go back to this idea of if we're going to dramatically change using AI, we have to have you as partners come along on this journey. We have to show you how the technology is actually working. I think that could be a combination of strategies.

Charles McKee:Explainability. The idea of being able to explain the methodology, the process, and that's a basis of trust building. Yes.

Eric Keeley:So you can— some of the strategies we're thinking about is, do we start small? And this, Marc, of course, you're the opposite. This is the opposite approach that you're taking, which I really respect, and I really like how you think about this. But we take a conservative airline that is, you know, very worried about these kinds of changes. We can create agents that can help make suggestions on what my changes might be, can communicate why these changes might impact and what is the impact.

Viktoriya Soubra:Yeah.

Eric Keeley:You know, maybe less disruptive time, maybe less cost, maybe less disruption to passengers, and then explaining that over time. This creates trust where then they start bypassing those explanations and just start trusting the system. We've even considered things like, you know, LLMs that will be able to explain or be more conversational to explain what's going on, or alerts in the system that can go rise up to the top and explain, hey, this is the impact.

Charles McKee:Right.

Eric Keeley:Broadly, because too many times the airline gets into these silos of, I'm an operations controller, I'm a crew scheduler, I care about airplanes, I care about crew. Those cannot exist in the AI world that Marc is very aptly describing. They've got to be combined, because we can't solve airplanes, airlines, or aircraft, and crew, and passengers unless we mix that understanding.

Charles McKee:So, okay, change management is— there's a A huge change management piece beyond the technology, right?

Eric Keeley:Yep.

Charles McKee:Do you want to respond to that?

Marc-Philippe Lumpé:If I may, of course, if you have a setup in an ops center that you're describing, I mean, you're anyways in deep trouble for everything because then the most basic processes don't work properly. So it might— but that's really— I mean, then you shouldn't even be thinking about AI or anything like that. You have to get your processes right first. Because otherwise nothing works.

Eric Keeley:I don't. I think it's a challenge that we're experiencing in the market with this technology is the hesitancy to use it. The question is, how do we move people into this next area of technology?

Charles McKee:I want to shift gears for a second. I'm going to put Max in the hot seat. We talked a little about the customer with Respect to your ops. But I want to talk more broadly about the commercial side of the airline business and the things that impact customers. You're in a very unique position. You're one of the leading PSS providers for a number of airlines. So you're in every day having conversations with airlines. You know, if we were doing this panel not so long ago, we might be talking about the death of the PNR, the evolution of reservation systems, etc. I have a question for you. Many questions for you, actually. But the first question is really, if we think about the extension of capabilities that artificial intelligence can potentially provide an airline, Mm-hmm. does that open up an opportunity to actually avoid ripping and replacing old technology with new technologies? Does it actually provide an extension of the useful life of the technologies that we as airlines rely on today? Or is this actually— should it be seen, should AI be seen as actually a great enabler, an accelerator to the inevitable evolution to the new technologies of offer, order, etc.? How are you thinking about that? And what are your clients saying to you?

Maximiliano Vicini:Okay, thank you, Charles. Good question. Well, first of all, for the ones who doesn't know HITIT, we are one of the leading PSS in the world. Again, we are the 3rd largest PSS after The 2 main players, Amadeus and Sabre. So, well, we are not only a PSS, we, we, we are more, more than that. We, we have a, a comprehensive airline solution suite that we call Crane, Crane Suite, that has the PSS, operational solutions, OCC, crew planning, loyalty systems, cargo, revenue accounting, cost accounting, and And more. So, well, the question here is not if we have to enable AI in a platform, but how to apply AI in a platform. So that's the first concept. From our side, we have created an AI framework. So we have been working with AI before ChatGPT. ChatGPT became a cool thing, you know. So our approach was first to launch what we call MVPs, minimum viable products, and test those things with AI enabled in different parts of our suite, you know. So going to the passenger side, you know, and first we have been able dynamic pricing and CDD dynamic pricing for personalization, you know. And, and well, in parallel of that, we enable NDC capabilities. We have the high levels of NDC capabilities, and we are certified in the Airline Retailing Maturity Index with the highest level of airline retailing maturity functionalities in our system for offer and order management. Settlement and delivery. And well, so that journey allows us now to have a product that has lots of personalization functionalities to give more options to the passenger. We have developed our— we work more on our external APIs rather than our standard ancillaries.

Eric Keeley:Okay.

Maximiliano Vicini:Unlimited ancillary capabilities and interlining and co-sharing capabilities for all our partner airlines in the world. So that is how we enable AI for these things— personalization, giving a better customer experience, for example, in our IBE platform, for the internet booking engine and so on. So that is on the passenger side. Of course, we also enable AI in other parts. Then we will talk about that in the operational side to optimise all the airline operations with, let's say, tail optimisers, crew optimisers, rostering and pairing optimisers, and so on.

Charles McKee:Let me ask a question. You're out there talking to airline clients every day. If you were to characterise the maturity level of those discussions? That is to say, when you think about your airline clients, are, number one, are airlines embracing artificial intelligence actively? Number two, are the discussions they're having with you as their technology vendor, are they ones that are really moving the ball forward? Is it a— where are we in the maturity curve in terms of artificial intelligence?

Maximiliano Vicini:Well, first, from HITIT's side, we already embrace artificial intelligence. So we have that. So then, okay, yeah, the challenge is the adoption of both technologies in different realities. Every single airline has their own reality, you know, their own business models. Even we can serve any business model today, and we already do that. But you have airlines that are in different contexts, you know. Let's talk about, for example, Latin America and the Caribbean, you know. Yeah. So it's not the same to run an airline in this region, and we cannot extrapolate that into another region or compare with another region. So what we offer to the airlines is the flexibility, the scalability, and all our AI-enabled functionalities for them. So the technology is there. But of course, there is another factor in the airline that is the, you know, the willing of adopting new technologies, you know, and that comes with sometimes with a cultural change inside the company, you know.

Charles McKee:Yep.

Maximiliano Vicini:So, and also what Eric said, discussed with you about the trust in the solution.

Viktoriya Soubra:Yeah.

Maximiliano Vicini:So another big challenge is to show to the airline that the AI investment, any AI investment today, comes with a real return of investment, you know, so means, and also to choose the right IT partner that can also monitor that AI behavior in the backend. You know, for not the AI to start hallucinating or, you know, and also have the necessary feasibility checks, what we already have in our, for example, in our operational solutions. So giving a warning to the operator that they need to fix something, let's say minimum connected times with the crew and those kinds of stuff. And so that is very important. But I mean, we made a survey among all our partners in the world. So we are operating now in all the continents and 50 countries. And technology, in that survey, technology was not the concern, but how the AI or, for example, offer and order management will will add value to the airline and how they can implement it. And in that sense, the IT partner is the key to help the airline to make this cultural change and, you know, sometimes go out from their comfort zone and adopt the technologies.

Viktoriya Soubra:Sure.

Charles McKee:So that's a really helpful kind of summation, I think, of where the— what we hear you say is that really the attitudes and the maturity really varies. I did actually just a quick survey of publicly announced AI projects here in Latin America amongst a sample of 5 airlines, and it was fascinating. The top one had 6 major public announcements about AI projects. That was followed by 4 for the next one, 4 for the third one, and 0 for the other 2. These are all fairly significant players in the region. So it's very interesting, I think, and we'll see adoption probably all reach a level of momentum.

Maximiliano Vicini:And going back to the PNR, the death of the PNR, you know, and what we have, if we have launches in new product called HITIT Oxygen that is a native offer and order management system, you know. And, and well, that will help to facilitate the transition from the old structure to the new offer and order management structure.

Charles McKee:Now we've got, we've got 8 minutes left. I want to do— I want to shift quickly, so I'm sorry about that, but I want to shift to make sure we cover not only the slide-out questions or Very much. But I actually want to really draw on your respective experience on deployment. You mentioned personalized pricing, and I'm going to use that as an example. There's a very well-publicized incident with a North American carrier that went out and said that a good proportion of their pricing modules are based on AI-driven personalized pricing. And there was a huge uproar about, well, isn't that discriminatory with the use of personal data? And it was a little bit of a tempest in a teapot. It wasn't really truly AI-driven personalised pricing, but it was, you know, they used the wrong terminology, I think. But nevertheless, that does beg the question, when we think about from an ethical standpoint or from a data protection standpoint, are there use cases of AI that airlines should not If I may, just from my personal experience, I have been lucky to be working with airlines, specifically in disruption area, for the past nearly 20 years.

Viktoriya Soubra:I have seen the change on how the adoption or acceptance of the fact that IROPS is going to be part of the operation It's only probably in the last 4 years where I've seen active development and build of specific departments to solve the IROPS.

Marc-Philippe Lumpé:Mm-hmm.

Viktoriya Soubra:When it comes to the adoption of AI that is related to the passenger, because we are looking after the passengers, we're managing the IROPS, and there's a lot of concern about privacy. personal data of the passenger.

Eric Keeley:Yes.

Viktoriya Soubra:Again, with iCoupon, the airline does not have to share any passenger information with us. We use a sequence number as a unique identifier, nothing else. However, what I have seen in terms of the change of the legal requirement as such, and I suppose that's probably because when we get the document, they are still not in full Understanding that we don't touch passenger data, but it is very black and white that there is no AI will be permitted in this operation. It's very restrictive and it's very limited as to what we can use to improve the passenger experience.

Charles McKee:I think that's a very important point. In fact, actually, it leads into one of our Slido questions here. We've got 3 Slido questions. So thank you for those that have provided these. The first question here is, what standards or certification frameworks are already available today to guide airlines in deploying AI safely and consistently? And when we talk about safely, in part that touches on what you're talking about, Viktoriya, which is the appropriate use of data. I guess I'd throw that open to the panel. Anyone like to comment on that? What standards or certification frameworks are already available to guide airlines in deploying AI safely?

Marc-Philippe Lumpé:I mean, you have, you have obviously data protection laws. For the rest, I mean, we are writing regulations as we develop this stuff in Canada because there is just nothing. And we have Transport Canada with us, and we are, as we're deploying the solutions, we are writing the regulations. So It's up to us, and that's probably also a bit the advantage when you're a technology leader. You have a big influence on how these things develop, how they're deployed, and how the regulatory framework afterwards works.

Charles McKee:Well, you know, it's interesting. In South America, Brazil has some very specific rules around the organisational use of AI. It's the only jurisdiction that does. There's some soft laws around data protection, but Brazil is the only one right now that has an AI Directed regulation. So I guess the question is, in this environment of not being overly regulated, does that actually enable innovation? Does that allow certain freedoms to operators like airlines to actually get out there and try things? And does it support the test and learn environment that we're in?

Marc-Philippe Lumpé:I mean, just from my point of view, the less regulation, the better it is. It's always the same thing. You get governments into this stuff and it stifles innovation.

Maximiliano Vicini:What I can say to that is we always follow IATA standards, ICAO standards, and also if in any single country when we deploy our, let's say, PSS or operational solutions on the PSS side, we always deal with GDPR. We are GDPR compliant for for the data security of the passengers and personal data, I mean. Then on the operational side, sometimes we must certify the solution with the civil aviation authorities. That happened with us in Brazil. It happened with us in other countries in the region.

Viktoriya Soubra:Okay.

Maximiliano Vicini:Sorry, and the rules, and it's important, is the rules that are in the system, that rules should of course be compatible with the government regulations, the union regulations in the case of the crew and everything. So that's how we control that.

Charles McKee:So I want to actually— the flip side of that is really How do airlines actually responsibly roll out AI, their practices, etc.? Eric, I want to talk to you. I want to ask you, in addition to working at CAE, you worked at a major aircraft manufacturer for 20 years. You worked very closely with a number of airline clients in the introduction of innovation. From your perspective, what are some of the— if you were to give a couple lessons on what does good, responsible rollout of AI look like in an airline? What are one or two lessons that you would share?

Eric Keeley:As a vendor, it's really hard because data is really addictive. Because with more data, I can solve more of your problems, and I want to solve your problems. That's the whole point of doing this. And so government regulation, of course, does hinder that to some degree. On the other hand, because you can do something doesn't mean that you should do something. And, you know, as you— what I didn't mention before was I was on the front line of bringing the iPad into the cockpit with navigation data. I was a part of that team. And we had the iPads that you have the ability to sense— we're doing experimentation now where it can actually sense, so sensitive that it can determine turbulence in the cockpit. And the idea was, oh my gosh, if we can do that, what if we can just tap into everybody's phones on the aircraft And we can tell a much higher level of turbulence and turbulence data, which will help safety. I think there's an obvious implication there that maybe that might be a bad idea.

Charles McKee:The outcome is good. The means is not so great.

Eric Keeley:Exactly. Yeah.

Charles McKee:We've got 23 seconds. I've got one question.

Viktoriya Soubra:All right?

Charles McKee:I'm going to go through. It's 5 seconds each. What do you think will be the biggest change in the way that customers shop, book, or consume travel in 5 years, 10 years because of AI? Let's start with you, Marc.

Marc-Philippe Lumpé:I think the customers are going to be much more guided probably in their decisions because it will be personalized. It will be very much up to what everybody wants. I think as long as customers don't feel manipulated but feel that they get actual additional value out of it and not just higher prices, Then this is going to be beneficial for everybody.

Charles McKee:We have the proverbial hook. I'm so sorry that we had one slide of questions we didn't get to. And I think we're out of time, though. So listen, thank you. This is a good discussion. Clearly, we could dedicate a whole CAPA conference on this topic. So thank you for your participation today.

Copyright policy: All transcripts on this site are the copyright of CAPA - Centre for Aviation. Our reproduction policy is as follows: you may quote up to 400 words of any transcript on the condition that you attribute the transcript to CAPA - Centre for Aviation and link to the original video page. All other use is prohibited. While we aim for 100% accuracy in the transcript, there may be some minor transcribing errors.