OTAs and Brands Forge AI Defenses as Travelers Hesitate

Online travel agencies and lodging brands are responding to agentic artificial intelligence by establishing external distribution partnerships, running parallel internal software experiments, and betting heavily on customer trust. While Silicon Valley developers promote autonomous agents, research shows travelers still rely on direct hotel sites, search engines, and proven booking engines to verify details and complete reservations.

How are major online travel agencies positioning against third-party AI agents?

Major online travel agencies are pairing distribution agreements with external tech giants while simultaneously building out internal tools. Speaking at the Skift Global Forum, Expedia CEO Ariane Gorin confirmed that Expedia is integrating with Meta's new AI agent, Muse [3]. Gorin stated that users can tell Muse where they are heading, allowing the agent to coordinate with Expedia to arrange hotels and intermediate travel details [3]. However, Gorin expressed clear skepticism that automated bots will replace standard booking paths, noting she does not believe a large portion of transactions will become end-to-end conversational experiences [3].

Booking Holdings is pursuing a portfolio model while dismissing the technical complexity of standalone conversational bots. According to Skift, Booking Holdings CEO Glenn Fogel addressed buzzworthy tools like Instinct by remarking that building an equivalent product would not be difficult [1]. Instead, Fogel emphasized that consumer trust, fraud protection, and handling sensitive credit card information are the decisive barriers to entry [1]. Skift reported that Booking Holdings is running five simultaneous AI approaches across its operations [6]. These encompass consumer-facing agents across Priceline, Agoda, and Booking.com, alongside investments in two native AI startups [6]. Fogel described this setup as a venture capital model where several tests can fail as long as one produces a significant win [6].

What does consumer behavior reveal about automated travel booking?

Consumer behavior reveals that travelers embrace generative software for top-of-funnel exploration but abandon automated agents when finalizing itineraries. A survey covered by Asian Hospitality showed that 86 percent of travelers have used AI for trip planning, yet only 5 percent follow through and book the exact suggestions provided [2]. Instead, prospective guests cross-reference AI suggestions with established channels: nearly six in 10 respondents search Google after getting an AI recommendation, half visit an official hotel website, and 40 percent compare prices across multiple travel websites [2].

Trish Nugent, senior vice president and head of public relations and public affairs at Mower, noted that being recommended by an algorithm is fundamentally different from being chosen, which elevates the importance of search visibility, official websites, reviews, and social channels [2]. The study showed that while 28 percent of respondents would begin planning a leisure trip with an AI platform—nearly matching the 30 percent who start on Google—consumer confidence remains tied to the complexity and price of the stay [2]. Travelers express openness to relying on algorithms for routine road trips, but confidence drops sharply when arranging flights, hotels, or luxury travel valued at $5,000 or more [2].

How are alternative accommodation platforms countering these changes?

Alternative accommodation platforms are planning full-scale proprietary concierge systems rather than relying purely on third-party channels. At the Skift Global Forum, Airbnb CEO Brian Chesky committed to launching an internal AI agent in 2027, describing the coming software interface as much richer than a conventional chatbot [4]. Chesky had outlined that guests will soon be able to submit broad queries, such as organizing a family vacation to Europe, with the platform assembling the itinerary [4].

This platform evolution coincides with shifting competitive boundaries across lodging sectors. Fogel noted that Airbnb's push into traditional hotels was inevitable [6]. He argued this move could ultimately benefit Booking Holdings by encouraging property hosts who previously relied on a single channel to list inventory across other platforms to secure demand [6].

Why does corporate infrastructure hold back hotel brand AI deployments?

Internal corporate fragmentation prevents hotel companies from realizing the commercial advantages of autonomous systems. Bret Taylor, founder of Sierra and chairman of the board at OpenAI, stated at the Skift Global Forum that artificial intelligence can already maintain a unified guest conversation across the reservation, pre-arrival, on-property, and post-stay phases [5]. Yet hospitality brands routinely undermine this capability by passing customers across disconnected internal departments [5]. Taylor warned that companies must avoid reflecting internal corporate divisions in customer-facing software, stressing that exposing an organizational chart to the guest is a severe operational mistake [5].

Taylor also observed that many operators mismanage artificial intelligence by treating it strictly as a cost-cutting measure for contact centers, directing the technology toward low-value customer service tickets [5]. He argued that the distinction between marketing and operations simply reflects who begins the interaction, advising executives to position autonomous agents as the front door for revenue, cross-selling, and guest support [5].

Where does commercial power sit between tech providers and inventory owners?

Commercial leverage remains concentrated with platforms that control customer service infrastructure, direct guest trust, and physical hospitality assets. Fogel pointed out that external AI referral traffic currently drives fewer than 1 percent of room nights at Booking.com [6]. While expecting transaction volume via conversational tools to increase over time, Fogel highlighted that generalized large language models lack customer support phone numbers when room disruptions occur [6]. He noted that Booking Holdings is designing Priceline's Penny agent to automatically resolve itinerary disruptions and book replacement inventory in real time [6].

As revealed by research from the NYU Tisch Center, RateGain, and HEDNA, more than half of hotels are already deploying or purchasing generative AI software [2]. For hotel owners and revenue directors, the immediate challenge is not ceding inventory margins to unproven AI startups. Because consumers systematically verify automated suggestions across Google, brand websites, and rate comparison engines [2], properties that maintain price integrity and robust digital direct channels can capture guests seeking confirmation before they commit funds.