Travel commerce is moving from discovery to decision-making as artificial intelligence (AI) agents control itineraries and transactions.
This transition is underpinned by a key axis of consumer trust. Nearly 25% of consumers say they would be comfortable trusting an AI agent to plan their trip, according to data from PYMNTS Intelligence. This is important because travel is a high-stakes, multi-transaction category that includes flights, hotels, ground transportation, and activities, often with limited flexibility and high costs.
As that comfort level increases, travel platforms are responding by deploying agents to plan itineraries, execute reservations, and manage trips across suppliers in real-time. As a result, travel commerce is moving from search-driven workflows to agent-driven execution.
From support to end-to-end execution
The first phase of AI adoption in travel focused on customer service and discovery. The chatbot answered questions, offered recommendations, and handled basic support. In the current phase, responsibility is transferred to autonomous systems designed to operate.
According to a report from CNBC, AI travel agents are moving far beyond the role of assistants to handle planning, booking, and disruption management autonomously. Agents can monitor prices, rebook flights, and adjust itineraries without direct user involvement, marking a shift from a reactive tool to a continuous decision-making system.
Expedia has deployed an AI-powered service agent designed to handle reservation changes, cancellations, and customer support issues in a single interaction. The company says the agent can solve problems that previously required multiple handoffs across search, service, and checkout. The result is a more continuous, agent-driven experience that reduces friction in the moments when travelers are most likely to abandon or escalate.
Advertisement: SCROLL TO CONTINUE
Trip.com's TripGenie extends this model to planning and booking. Agents generate comprehensive itineraries across flights, hotels, and activities, dynamically adapting recommendations, and complete reservations directly within the same flow. Rather than returning search results, TripGenie constructs and executes trips based on the traveler's intent, budget, and timing constraints, shifting the cognitive load from the user to the system.
Delegation is also expanding through devices. Amazon's Alexa Plus now supports voice-based travel booking integration, allowing users to initiate and complete reservations through voice commands. Alexa conveys context across plans, reservations, and changes, reducing repetitive input and empowering agents as a system of record rather than a one-time interface.
This pattern is also emerging in managed and corporate travel. Amadeus, Microsoft and Accenture have collaborated on a travel planning agent available to users of Amadeus' Cytric Easy business travel platform. Employees can chat with a natural language agent within Microsoft Teams to plan and book travel, replacing traditional sequential searches with conversational, agent-driven workflows. Agents coordinate policy compliance, availability, and reservations in one interaction.
As agents increasingly become the gateway to travel planning and booking, giving consumers control over where and how they begin their journeys has become a strategic priority for the largest online travel platforms. According to the Wall Street Journal, Booking Holdings is positioning AI as a discovery channel rather than a threat to its business as chatbots change the way travelers plan their trips.
CEO Glenn Fogel told WSJ that travelers are “starting to change how they begin their travel inspiration, planning, and discovery. But we want to be there when they start their journey.” Vogel said that in the same way that Google has historically driven demand, Booking sees AI as another gateway to its platform and expects scale and global reach to continue to dominate as AI providers decide which travel platforms to list their chatbots on.
Beyond platforms
The economics of travel platforms are changing as agent systems take control of execution.
Historically, online travel agencies have optimized for clicks, conversions, and upsells. Agentic AI restructures the product around the outcome: whether trips are successfully planned, booked, and managed with minimal friction. The interface is less important than the agent's ability to complete the job end-to-end.
This shift is visible beyond consumer platforms. GDS Genie is an AI-powered orchestration layer that coordinates inventory across global distribution systems. Rather than ranking results for users to compare, the system synchronizes availability, rules, and pricing across suppliers, allowing agents to make actionable decisions rather than presenting piecemeal options.
The same orchestration logic is now being extended to hospitality operations. Google Cloud and Radisson's partnership shows how AI connects personalization, reservations, and operations. For Radisson, AI models help tailor offers, streamline guest interactions, and connect front-end booking decisions directly to downstream service delivery, reducing operational friction and improving consistency across the guest experience.
