Hoteliers are grappling with infrastructure and data fragmentation as companies revamp their tech stacks for the artificial intelligence (AI) era.
Stakeholders are taking different approaches to AI investment, with some opting for targeted AI deployments and others pursuing broader modernization efforts.
“Both approaches exist in the market right now,” said Tan Bee Leng, chief commercial officer for The Ascott Limited. “Some players are rebuilding from scratch, others are layering on top of what they have.”
How hotels are modernizing the tech stack
Eighty-two percent of hotels planned to increase AI usage in the next 12 months according to March 2026 research from Canary Technologies. With varying legacy architectures in place and no blueprint for AI evolution, hoteliers’ execution philosophies vary significantly by brand footprint and business model.
“The differentiator for hospitality brands will be how effectively they integrate AI across discovery, booking, operations and distribution,” said Doug Lange, VP of IT strategy at Choice Hotels. “Those that treat AI as core infrastructure and not an add-on will be among the best positioned to compete.”
Lange said Choice Hotels is choosing to embed AI across the entire enterprise as opposed to layering point solutions on top of systems that are already disjointed.
The business is moving its core functions to cloud-based platforms such as Amazon Web Services’ cloud infrastructure to weave intelligence through the enterprise.
“Many legacy systems were designed for a pre‑AI world, which was optimized for static workflows rather than real-time learning and intelligence,” Lange said. “Our cloud-first approach has positioned us really well to take advantage of the capabilities AI brings to the forefront.”
The shift to the cloud has enabled improved security, flexibility and scalability, Lange added.
“By standardizing platforms and operating on [a] shared cloud foundation, we’re creating a more unified, intelligent enterprise, where insights and capabilities can be reused across teams and functions instead of rebuilt repeatedly,” Lange said.
Minor Hotels is undergoing a ground-up refresh, Ian Di Tullio, chief commercial officer for Minor said, noting that most hotel stacks haven’t been designed end-to-end.
“PMS [property management system], CRS [central reservation system], CRM [customer relationship management platform]—pick your acronym—different systems have been added at different times, often across brands and regions, with each solving a specific problem,” Di Tullio said. “You end up with plenty of data, but also a lot of fragmentation and very little real-time visibility.”
Adding another layer wasn’t a viable fix for Minor, Di Tullio said. Starting again allows a full foundational rebuild.
The focus is on a unified layer that everything else connects to, Di Tullio said.
“We talk about the underlying technology but this is really about a shift in mindset: from systems talking to each other to systems working off the same source of truth in real time,” Di Tullio said.
Meanwhile, Ascott, which operates with an asset-light strategy, is navigating the shift by building a multi-tier architecture to support AI and human interactions.
“Our view is that competitiveness in AI does not come from replacing systems alone,” Tan said. “It comes from how well your data and interfaces are structured for AI to actually work with [them].”
The foundation includes core systems that cover inventory, pricing and transactions, she said. On top of that is an integration and orchestration layer to standardize how data and systems are accessed and shared—which allows the company’s AI to work reliably. Finally, an AI interface layer sits on top, including Cubby, Ascott’s conversational assistant, and other external AI platforms.
The key principle is one interface that AI channels can connect with, Tan added.
“When that is in place, we can scale to new use cases without rebuilding the foundation each time,” Tan said. “It is less about any one technology and more about how the architecture holds together.”
The data fragmentation hurdle
Stakeholders agreed data fragmentation is one of the largest hurdles to scaling AI in hospitality, with Di Tullio flagging it as “the main constraint” on AI.
Since hospitality data sits across different layers, is often inconsistent and isn’t synchronized, hotels are limited on what they can actually do with data in real time, Di Tullio said.
For some hotels, the operating model plays into all of these decisions.
What a hotel brand can do with its data to prepare for AI comes down to its operating model, according to Di Tullio.
“The ability to standardize data at this depth depends on owning enough of the operating environment to enforce that standardization,” Di Tullio said. “A franchise-heavy group with hundreds of independent operators can’t unify a guest profile by issuing a memo. We can—because we own or lease the majority of our portfolio. The operating model is what makes the data model possible.”
Di Tullio said Minor is addressing the data layer first and foremost. The company is seeking to create a “unified, persistent guest identity,” and in doing so is standardizing how data is structured across its stack.
“We’re not replacing every system. But we are reducing duplication and making sure those systems operate against a consistent data model,” Di Tullio said.
Tan said that Ascott is addressing fragmentation at the data and integration level.
The company’s goal is to have a consistent, governed layer of data that works as “a single source of truth” for content, pricing and availability, she said. Systems will be connected by an integration layer that guarantees reliable data flow in real time.
AI initiatives will sit on top of that layer, with the idea that these deployments will deliver a better experience because the foundation is clean and consistent, Tan noted. Ascott is not reducing systems for the sake of cutting back but to ensure the systems operate as one unit.
While approaches vary, the foundational constraint remains the same. The hotel sector has to address legacy tech to fully leverage AI.