Hype versus reality: How fast is generative AI changing the travel industry?

The hype around artificial intelligence and, more specifically generative artificial intelligence, has been palpable over the last nearly two years.

Airbnb CEO Brian Chesky is among leaders who have touted how their companies might evolve thanks to generative AI. But the change is yet to come – a fact Chesky acknowledged at an event he spoke at in September.

“I don’t like to make promises, and not deliver on them … but, well, I’m not going to deliver on the promise of AI, and no one will,” said Chesky. “I think AI is going to change the world much more than anybody realizes. I also think it’s going to take way longer than anyone realizes. This decade, things aren’t going to change as much as people think, and next decade, things are going to change a lot more.”

Have expectations set by industry leaders around AI fallen flat? Is generative AI-powered change happening as quickly as anticipated? These questions feel relevant, especially with the two-year anniversary of ChatGPT’s launch approaching next month.

It’s easy to believe that change will be rapid, but AI can only move as fast as humans are able to move with it.

Mike Coletta, Phocuswright

Mike Coletta, senior manager of research and innovation at Phocuswright, said there is a great dichotomy around AI that makes it difficult to wrap one’s head around the technology’s evolution.

“On the one hand, it’s by far the fastest-evolving technology in history,” Coletta said. “On the other hand, humans at large evolve at a set pace, both in our understanding of technology and our behaviors with it. So, it’s easy to believe that change will be rapid, but AI can only move as fast as humans are able to move with it.”

PhocusWire reached out to travel industry leaders to gauge their opinions on the pace at which AI is changing the travel industry – some believe Chesky’s point is valid, others believe an AI-powered future is here.

Responses were received via email and have been edited and condensed.

Sundar Narasimhan, senior vice president, president of Sabre Labs and product strategy, Sabre:

There is a lot of capital being invested in AI and soon investors will begin to want to see returns for that investment. My view is that – compared with other emerging capabilities like blockchain or VR/AR – AI (especially LLM / generative AI) is more tech-ready and embeddable within products and use-cases will begin to emerge across the industry. Product-market fit might be harder to come by, but I think that we will see more compelling use cases emerging in the next few years.

Amy Read, vice president of innovation, Sabre:

AI has been around for a while, and although the hype has surged since the launch of consumer-facing GenAI chatbots, leading to massive investments in the field, it’s important to keep a balanced perspective. Do we believe AI is transformative? Absolutely. But I also think it will take time to fully realize its potential in the market. It reminds me of the dot-com boom and bust — everyone was eager to sprint before we even learned to walk, and the industry and technology needed time to catch up.

There is a lot of investment in AI, but it’s crucial to remember that these models will only be as effective as the data that drives them. What’s important now from an industry perspective is building solid foundations—organizing internal resources and data to be future-proof. If we can then apply AI in a meaningful way to address what guests need by simplifying and personalizing their travel experiences, one use case at a time, then we will accelerate its adoption.

Line Crieloue, vice president of image and influence, Accor:

We need to differentiate AI from GenAI. AI has been around for decades now and is already bringing disruptions … like revenue management or personalization for instance.

GenAI is much newer. While it is moving at a fast pace, there [are] still a lot of improvements yet to come in particular around reducing hallucinations and training models that perform in multilingual and multicultural environments. Yet we do observe already a proliferation of impactful use cases in particular in domains where the risk of error has a low impact. This includes for instance use cases for internal productivity or assisted sales. What will for sure take time is the improvement of these models and therefore the acceleration of fully automated use cases such as customer-facing features or marketing related topics. When these accelerate, it will for sure disrupt our sector.

Sanjay Mohan, group chief technology officer, MakeMyTrip:

I’ll agree that the GenAI hype has been drummed up quite aggressively – a bit too aggressively I’d say. … Reality will eventually catch up with the hype, and will meet or beat most predictions, as I can see. … We’re still very, very early in this game. The potential is huge, and this GenAI tech is here for the long haul, so let’s not judge too early. It’ll definitely deliver against the promises made.

The obvious use-cases have been built and shipped. We’ve shipped features that utilize content analysis and summarization in our hotels funnel. We’ve also shipped out some interesting features employing voice in regional Indian languages that effectively assist the end users while they’re in our product flows. This is huge for a country like India, where English is limited to tier-1 and tier-2 cities mostly. The GenAI tech is fostering inclusivity, if you view from the language lens. The call center agents are using GenAI-enabled features very effectively. Our employees are using productivity tools for coding as well as analysis of data on a daily basis. So the gains are visible all across – externally and internally. The GenAI technology is already delivering and creating impact. We, ourselves, are getting ambitious in taking up more complex features where GenAI can create significant impact in easing the lives of the travelers.

The technology, itself, is evolving at a very rapid pace – on all fronts. The capabilities are getting better every month, the costs are coming down, and the accuracy/latency is getting better. If you look at the larger landscape of the pace of underlying technology, and the ambitions of the application providers (including OTAs), I am very confident that this decade will end radically different than what any of us could have imagined a few years ago (before GenAI arrived on the scene). The tech is here to stay, and will deliver on all promises – mostly within this decade, as I see it. Beyond that, there will [be] a new set of dreams and imaginations – and that’s for the next decade.

Adam Harris, CEO, CloudBeds:

AI isn’t advancing in our industry as quickly as it could, even though hospitality was one of the early adopters of AI, particularly with revenue management systems, which have been in use for decades to optimize pricing and inventory. More recently, AI chatbots have also become common, enhancing guest experiences by automating tasks like answering FAQs and streamlining bookings.

However, while these point solutions are powerful, they don’t fulfill the broader promise of AI. Each tool optimizes just one department — revenue, guest services, or operations — leading to isolated gains without solving inefficiencies across the whole organization.

Running a successful hotel involves many interconnected systems, and optimizing just one area while leaving others untouched still results in inefficiencies. As Bill Gates has said, the impact of AI should be as transformative as the creation of the internet or the smartphone —both of which succeeded because they democratized data and unlocked possibilities across multiple fields. For AI to truly revolutionize hospitality, it needs to break down the silos that exist between departments like revenue, marketing and guest experience.

This fragmentation of data is the main roadblock. AI can only be as effective as the data it is fed, and in many hotels, different departments work with separate systems that don’t communicate with each other. Until a solution is developed that captures and unifies all these critical data points, the impact of AI will remain limited to narrow functions. … Our connected, open platform captures billions of data points from across the industry, allowing us to build a sophisticated Causal AI model that can predict demand and optimize operations across the entire organization. This unified approach unlocks the true potential of AI, helping hotels drive a comprehensive commercial strategy rather than optimizing isolated KPIs.

So while AI is making strides in hospitality, its transformative potential will only be realized once the industry moves beyond departmental silos and adopts platforms capable of integrating data across the entire hotel ecosystem. This holistic approach is essential to achieving the sweeping changes the industry has been waiting for.

Alex Bainbridge, CEO, Autoura:

I believe the change is here now, but it’s hard getting AI into the hands of consumers as the business models that AI enables don’t really work with how the sectors currently work.

For example, AI trip planning: Now you have 10,000 itineraries in a city rather than 100. How do you commercially operate that? And if you can’t operate them, how do you monetize those itineraries beyond attraction booking (which means you have to compete head on with the OTAs on day 1)? 

Timings are incredibly difficult to judge here….. and tech is always slow [at the industry wide scale] …… but I think maybe two to three years for significant change (US / EU) rather than end of decade.

Andres Martinez Artal, founder and CEO, Speakspots.com:

While we’ve seen widespread integration of “smart” chatbots by travel startups and incumbents, their limitations are evident. They struggle with ‘proactive’ tasks such as customizing genuinely unique trips. They also struggle with last-minute travel disruptions due to limited breadth of real-time data.

In order for AI to transform, rather than simply enhance the travel planning and experience, there are barriers that will take years to overcome. Data silos, such as hotel brands, airlines and OTA’s private APIs “by invitation only” for real-time data, limit AI agents’ access to quality data. The public internet is not enough.

Better integration of real-time data from IoT devices and wide adoption of voice assistants on wearables could improve this. “Anonymized” conversation data is a great real-time data source. We also need tighter integration across booking, payment, and authentication systems. Hotels, with direct access to travelers and, most importantly, leisure providers are positioned to lead AI advancements at the destination level.

Having the traveler’s personal AI agent exchange information with the hotel’s AI agent will allow hotels to customize the experience further without need for any receptionist-customer conversation. First, we need massive adoption of AI voice assistants by society. The issue is not for hotels to adopt AI, many of them are already actively adopting AI where they can. The gap is mobile related to the purchase and replacement cycle of mobile technology. People only buy a new mobile phone every two to five years. So even once “new-generation” smartphones are on the market with integrated AI voice technology across the board, we’ll still be looking at another two to five years before this tech becomes the mainstream in mobile. 

Ilan Twig, CTO and co-founder, Navan:

While the next decade promises revolutionary changes driven by generative AI, I’m thrilled to say we’re witnessing transformative impacts today. The debut of ChatGPT ignited global excitement — and yes, some overblown hype. But forward-thinking companies tackling AI’s challenges head-on are already reaping remarkable rewards.

…We created Ava, an AI-powered chatbot trained to think like a seasoned problem-solver. The journey wasn’t smooth — we faced hallucinations, biases, and security challenges — but these obstacles fueled our innovation. We developed Navan Cognition, a game-changing process that combines the best elements of various large language models while eliminating error-prone functions. The results? Spectacular. Ava now resolves more than 50% of support requests, rivaling human agents in satisfaction scores and saving us $1 million monthly.

Looking ahead, we envision a revolutionary, proactive UI powered by Navan Cognition within five years. Picture intuitive applications that show users exactly what they need, when they need it. We’re betting big on chat becoming the primary interaction mode, with increasingly sophisticated voice capabilities integrated into AI bots.

The AI future isn’t just coming — it’s here, and it’s more exciting than we ever imagined. Success in this AI gold rush demands diving in early, addressing obstacles, and continuously pushing boundaries. The transformation is happening now, and it’s just the beginning of an exhilarating journey.

Mike Coletta, senior manager of research and innovation, Phocuswright:

Skeptics who point to the excitement and disappointment curves of technologies like virtual reality or Web3 should be under no illusions about AI … the transformative capabilities of AI are fundamentally superior. Thus, those who can move fastest, the early adopters, will have a more distinct advantage than ever. We will see further step changes in the capability of the underlying models in rapid succession, and it will be critical for companies to be in a position to adopt them as they arrive. While there is always risk in adopting a new technology, the rewards with AI are clear.

In travel, these rewards are already being reaped, as we recently detailed in From Buzzword to Bottom Line: Keeping Pace With Gen AI in Travel. And while the there is much uncertainty about how AI will impact industry and the world in the medium to long-term, we’ve outlined how travel companies should position themselves for the rewards of the future in True Automated AI in Travel Is Coming, which includes the anticipated impact of autonomous AI agents assisting both consumers with shopping and booking and businesses with automation and efficiency. In essence, it appears likely that AI will upend everything in travel from user interfaces to distribution to the customer experience to the workforce.

Much like with the internet, mobile phones and social media, change seems to be happening slowly for a while and then it appears to happen all at once. Fundamental change will indeed take many years, as humans learn to adapt. But it’s already late to be getting started.