"It’s not a software issue, it’s an ethics issue": IBS Software CEO Somit Goyal on AI, Pricing, and Discrimination
Is AI a bubble? "Yes, in the 2-3 year view," says IBS CEO Somit Goyal at the ALTA forum. He argues its true value is long-term, enabling personalization, but warns airlines must navigate the critical ethical line between a good offer and "discrimination."
/https://aviacionlinecdn.eleco.com.ar/media/2025/10/somit_goyal_ceo_ibs_software.jpg)
Is the artificial intelligence revolution a massive bubble or the future of aviation? For Somit Goyal, CEO of IBS Software, it’s both—and it’s "the fastest growing technology I’ve ever seen." At the ALTA AGM & Airline Leaders Forum 2025 in Lima, we sat down with Goyal to cut through the hype and discuss what AI means for an industry still running on 40-year-old systems.
The conversation quickly zeroed in on the critical balance between powerful personalization and the ethical peril of discrimination, exploring why he believes the industry's future isn't just about software, but about "ethical development."
Pablo Diaz, Aviacionline: First of all, thank you for your time. Is this your first time here in Peru?
Somit Goyal, CEO of IBS Software: Not in the region. I've been to Latin America before, but it's my first time in Peru.
Pablo Diaz: How do you like it so far?
Somit Goyal: Wonderful. I love whatever I've seen so far. It's a beautiful country, great weather, great people.
Pablo Diaz: My first question is about artificial intelligence. There is a revolution happening now, but do you think we are in an AI bubble?
Somit Goyal: It's a fantastic question. I think if you think about how much money is flowing into AI, if you look at the hundreds of billions of dollars that are being pushed into making AI data centers and everything, you might think that it's a bubble.
It really depends on what lens you put on it. If you look at the pure amount of money that is flowing in, have the companies really received the return on that investment? No, of course, that's not happening.
However, what you should see is how quickly the technology is improving. Every month, every week, there's new technology, new capabilities coming. What you can do now, you couldn't do four months ago. This is the fastest-growing technology I've ever seen in my career, at least. So you can imagine what will be possible to do through it two years from now, if you take the same trajectory.
I think that if you take a two-to-three-year view of it, then yes, it's a bubble. But if you take a 10 to 15-year view of it, then I think it's not a bubble, because I think all these investments will finally bear fruit starting, let's say, three to five years from now, because the technology will just become better and better over time.
Pablo Diaz: So, it’s not FOMO, it’s more about "coming early to the party."
Somit Goyal: Correct.
Pablo Diaz: Following up on that, I was discussing revenue management with some folks here. That's a heavily perfected system that has been optimized since the '80s. What real value is AI adding to something already so advanced?
Somit Goyal: I think what we can do is, if you really look at the kind of data airlines have, if you look at the type of data loyalty systems have, it's a humongous amount of data. And then you ask yourself the question, have the airlines really done anything with it in the last 10, 20 years?
The answer is no. I mean, there's some basic stuff, but if you think about the retail systems of the future, and how you could marry up the loyalty systems with potential retail systems, what kind of scenarios can you create?
The example that I use is very pertinent to this conference. I was talking to one of the CEOs last night, and they said, "Hey, all the attendees that were coming to this conference were asking us, 'Can you actually create a three-day package for us for Peru?' Because they're doing the conference, they might as well stay for three more days, visit different places." This is the classic scenario where you should be able to see someone is coming to Peru and they're coming for a conference.
But what if you offer them something? And you also know something else about them because they have been buying your products from the loyalty system. After all, they were spending miles. So, you can truly customize the offer for them. And that's actually really good for them because it's highly personalized. It's good for the airlines because it's a better revenue opportunity.
So that's an example of a classic scenario where technology is available today to support it. But today's systems don't work that way. These are like 30, 40-year-old systems. The messaging system is EDIFACT, and all of that stuff. But as these things change, AI will enable these scenarios, which were not possible so far.
Pablo Diaz: In the software industry, we always had the motto, "Don't change anything that is working," because the risk often outweighs the benefit. Is this new wave of technology mature enough for companies to take the risk of upgrading their core systems?
Somit Goyal: I think so. And the reason... because I personally think not everything is about AI, to be honest. If you think about a lot of the scenarios, AI didn't even need to exist for those scenarios to exist today. The problem is in the airline commercial [space], the dynamic is it's been a duopoly for the last 30, 40 years.
There have been two players in the industry. And whenever you have that sort of market dynamic in any industry, forget airlines, it always stops innovation from happening. And I say that as someone who comes from Microsoft, right? It's just a duopoly.
What I'm saying is that a lot of these things... I'll give you another example. Last year, my wife booked a ticket from Seattle to Singapore. And because she booked a year in advance on an economy class ticket, she got a dirt-cheap rate. We got that ticket for $300.
Now, what happened was that, like two months after that, the need went away, so she didn't want to travel anymore. But because it was a non-refundable ticket, I cannot sell that back to Singapore. I cannot return it. While I can guarantee you that Singapore Airlines can now sell that ticket for at least three or four times that price.
Pablo Diaz: And the ancillaries that your wife wasn't taking.
Somit Goyal: Exactly. So now, Singapore Airlines is losing money, which they can easily make if the capability existed in the system. And it requires no AI. It's just like a reverse auction kind of thing. So that's like a classic scenario where the benefit, coming back to your question, the reason I gave that example was, some of the benefits are so obvious, but not supported by current systems.
Pablo Diaz: I was also thinking about personalization. With this technology, you can find the perfect offering for a customer. But as I asked in the panel, isn't there a risk of finding the "perfect price" for everyone, and couldn't that become a liability?
Somit Goyal: No, I don't think there's a thing called “perfect price.” I don't think you can... this was my point in the panel around the distinction between personalization versus discrimination. Because if you can hyper-personalize, you can go into the discrimination world, which is really bad. So like, if my mom has cancer, I just found out that my mom has cancer, for example, and I have to buy a ticket for the next 24 hours, and you somehow know because on my Facebook, I said my mom has cancer, so you know that...
Pablo Diaz: You have cookies, you have seen treatments, or anything.
Somit Goyal: And you can triple the price for me. That might be the perfect price from an airline perspective, but you don't want to do that. But if you stay in the personalization area and do not cross over into the discrimination area, which means you only give me an offer which is for my benefit, not to harm me... then I think... so if you go back to that Peru example where you're offering me a package, then you're probably giving me the right package at the right time at the right price.
And I have an opportunity to participate in the surplus. It's not going away from me completely. I'm pulling some of it to my side as well, as a consumer.
Pablo Diaz: So, it is not a software issue; it’s an ethics issue. The constraints need to be ethical.
Somit Goyal: Absolutely. I mean, technology... It's software. You can do anything. Anything's possible. It's about ethical development.
Pablo Diaz: Just three quick questions I ask in every interview: One book, one movie, and one song.
Somit Goyal: Oh man, this is... I'm a musician, so this is very hard.
Pablo Diaz: Really? What do you play?
Somit Goyal: I play the drums. I'm a drummer. Books... I think the one from Walter Isaacson, [who wrote] the Steve Jobs biography. So that was a very impactful book for me. I love... I'm not a fiction reader. I'm a non-fiction kind of person. So Steve Jobs' biography. There's also... this is a weird book, it's on Andre Agassi, the tennis player. He wrote an autobiography called Open.
Pablo Diaz: It's a great book. It delves into the mind of a fierce competitor.
Somit Goyal: Amazing book. The amount of authenticity and genuineness... Yeah. That's actually one of my favorite books. So I would say those two I'll call out. Favorite movie? Man, there are so many.
Pablo Diaz: Just the one that, if it is on TV, you have to stop and watch it.
Somit Goyal: I'd probably say Forrest Gump.
Pablo Diaz: All right. Good choice. And songs?
Somit Goyal: Anything from Led Zeppelin. That's my favorite band.
Pablo Diaz, Aviacionline: Oh, yes. Well, if you're a drummer, it's kind of obvious.
Is there anything I didn't ask?
Somit Goyal, CEO of IBS Software: No, this was a fantastic conversation. Thank you for approaching me and asking for the conversation.
Para comentar, debés estar registradoPor favor, iniciá sesión