Ascott Unveils Agentic Travel Commerce Strategy: Accenture, Amadeus and EHL Partnership Builds AI-Agent-Ready Booking Infrastructure

Hospitality's first formal "agentic commerce" declaration

Ascott advances AI-ready strategy to support agentic commerce

AI in hospitality drives Ascott strategy with Accenture, Amadeus and EHL, enabling agentic commerce and AI-led travel booking.

www.traveldailynews.asia

Singapore-based Ascott Limited has made a move that stands out in hospitality. The lodging arm of CapitaLand Investment runs more than 1,000 properties across 14 brands and 230 cities, making it the world's second-largest serviced apartment operator. Between April and May 2026, that company announced a sweeping AI-ready infrastructure investment combined with a three-way partnership spanning Accenture, Amadeus and EHL Hospitality Business School.

What makes this more than another "AI chatbot rollout" is the headline framing. Ascott is squaring up to agentic commerce, the emerging model in which AI agents search, compare and ultimately book on the traveller's behalf, completing transactions end-to-end without human web navigation.

Kevin Goh, CEO of Ascott, made the strategic posture explicit: "Distribution shifts, labour pressures and rising guest expectations are reshaping hospitality. Instead of waiting to see how agentic AI plays out in travel, we are building the infrastructure to shape how it does." Choosing to design the rails rather than ride them is what separates this from a typical IT modernisation programme.

Why a serviced-apartment giant moved now

Three pressures converged on hospitality over the past 12 months and explain the timing.

The first is the mainstreaming of AI-agent-mediated travel search. At its October 2025 DevDay, OpenAI launched Apps in ChatGPT with Expedia and Booking.com as initial travel partners. Travellers can now describe dates and preferences inside ChatGPT and receive live availability, prices and images without leaving the chat surface (OpenAI, PhocusWire).

The second pressure is Google's entry. Marriott CEO Anthony Capuano announced a Google AI Mode direct-booking integration in February 2026, letting users complete reservations in natural language without ever touching marriott.com. Both Marriott and Hilton have begun listing AI platforms as a new disclosed risk to direct bookings in their 10-K filings, signalling that AI agents are now treated as both threat and opportunity at the board level (Skift).

The third is the distribution-standard shift toward attribute-based shopping. Traditional hotel inventory has been expressed as room-type plus rate-plan combinations, but AI agents demand richer attributes (view, bed configuration, desk, long-stay amenities). Marriott migrated to the Amadeus Central Reservations System in late 2025, Accor followed, and Ascott now joins the same standard, putting it in lockstep with the world's leading lodging chains (Amadeus, PhocusWire on Accor).

A three-layer division of labour

The three partners Ascott chose are not interchangeable. They map cleanly onto distinct layers, and the design philosophy is to rewire architecture, distribution and people simultaneously rather than sequentially.

Accenture Song carries the architecture layer, redesigning the digital backbone that connects reservations, property management, CRM and loyalty. Emily Weiss, Accenture's Global Travel Industry Lead, framed the bet bluntly: agentic commerce is the largest commerce shift in 20 years, a technology overlay is not enough, and platforms need to be re-architected so machines can transact with brands. The concrete deliverable is moving Cubby from travel companion to booking agent.

Amadeus owns the distribution layer. Ascott's adoption of the Amadeus Central Reservations System unlocks API-first, attribute-based inventory, faster promotion activation, consistent pricing logic and tighter matching between guest preferences and available rooms. Paul Wilson, VP Asia Pacific Hospitality at Amadeus, characterised attribute-based shopping as "where distribution is heading", reframing ACRS not as a back-office upgrade but as the machine-readable shelf that AI agents will browse.

EHL Hospitality Business School handles people. Through Ascott's Global Brand Academy, EHL is building internal certified trainers, starting with the Ascott brand and rolling out to Oakwood, Citadines and beyond. The thesis is that judgment and warmth are the human edge that AI cannot replicate.

What Cubby's role change really means

Cubby deserves its own beat. Since 2023 the in-house digital concierge has handled more than 900,000 guest enquiries on DiscoverASR.com, autonomously resolving routine interactions and contributing to bookings. The new chapter promotes Cubby from chatbot to a transaction-capable agent.

Tan Bee Leng, Chief Commercial Officer of Ascott, summarised the shift: "In an agent-led travel ecosystem, Ascott properties will have to be visible where the real decisions are made — inside algorithms. Brand and property information must become machine-readable and optimised for generative engines."

The implication is not a UI refresh. It is the recognition that the brand's first touchpoint is now an algorithm, not a screen. Cubby is being positioned to interoperate with external agents (ChatGPT, Gemini, Claude) while the Ascott Star Rewards programme is required to deliver the same loyalty experience whether the booking originates on a brand site or through an agent.

How Ascott compares to the wider field

Placed in the broader race, Ascott's announcement is not a niche serviced-apartment story. It is a chapter in the industry-wide rebuild of distribution for the AI era.

Expedia and Booking.com, as OTAs, are pursuing both ChatGPT-app embedment and tighter funnels back to their owned platforms (Expedia in ChatGPT, PYMNTS). Marriott and Hilton are pushing the brand-direct route, landing the booking on their own engines even when discovery happens on an AI surface (TheStreet on Hilton AI Planner).

Ascott has clearly chosen the brand-direct camp, but with a twist: it is also building Cubby as the hub through which external agents plug in. That is a brand-sovereignty model for the agent era, executed at global scale.

Implications for commerce and lodging operators

For operators outside the Ascott orbit, three takeaways carry over.

First, attribute-isation and machine readability of product data are no longer optional. AI agents reason over structured attributes rather than free-form copy. Operators stuck with thin "room type plus rate" data risk being filtered out before agent ranking even begins. The "Unified Digital Shelf" concept that surfaced around Marriott-Google applies equally to non-lodging commerce (TRAVHOTECH).

Second, loyalty design has to be reopened. Ascott repeats the principle that the same personalisation and rewards must flow whether the guest books on the brand site or through an external agent. That raises hard questions about how to pass member identity to an agent and which party guarantees points accrual.

Third, owning a first-party agent is becoming a strategic hedge. Cubby is the levered bet: by housing the orchestration layer in-house, Ascott reduces dependence on external AI platforms. The Accenture engagement is, in effect, the architecture cost of keeping the agent in-house.

Conclusion

Ascott's announcement marks hospitality's transition from "what can AI do for us" to "how do we adapt commerce when AI agents are the default". The Accenture-Amadeus-EHL trio shows that the answer requires moving architecture, distribution and people in parallel.

The signal to watch next is whether other operators will follow with both a machine-readable shelf and a first-party agent. In the next quarter, count how often these two ideas appear in earnings calls and tech-partnership announcements — that is the cleanest gauge of who is genuinely building for agentic commerce.