Key Takeaways
- At Explore 2026, its annual partner conference held in May 2026, Expedia Group positioned AI as a growth engine, introducing three Partner Central agents, an in-development B2B AI toolkit, and a response to B2A, discovery aimed at AI agents
- The company's research found that 68 percent of travelers choose trusted brands over AI chatbots when booking, a marketplace quantifying for itself where AI commerce stands today: discovery is moving to AI while checkout has not
- For retail and e-commerce operators, the lessons are that machine-readable content is now the condition for being discovered by AI, and that channel and trust design should assume the asymmetry between discovery and checkout
A Marketplace in Its 30th Year Rebuilds on AI
Expedia Group Explore 2026 shows AI reshaping travel industry workflows, boosting bookings, marketing, operations and partner performance globally.
www.travelandtourworld.comExpedia Group held Explore 2026, its annual partner conference, in Las Vegas on May 19 and 20, 2026. More than 2,000 people from over 1,000 partner companies gathered at an event marking the company's 30th anniversary, and the message it chose to lead with was that AI is no longer an experiment but a growth engine. From inspiration and planning through booking, operations, disruption support, and marketing, the company laid out a plan to re-embed AI as the foundation of the business rather than a bolt-on feature.
This article reads the announcements not as travel-industry news but as a case study in AI commerce. AI commerce means selling in which generative AI and conversational agents enter the journey from product discovery to purchase. The stage within it where AI agents carry out comparison and transactions on the user's behalf is called agentic commerce. Expedia is an online travel agency (OTA) dealing in hotels and flights, but at its core it is a marketplace that aggregates inventory and connects it to demand.
That is precisely why the conference matters to retail and e-commerce readers. If the entry point of search moves to ChatGPT or Google's AI, intermediaries like OTAs are the first to feel the pressure of disintermediation. A business sitting in structurally the same seat as marketplace e-commerce platforms and price-comparison sites showed how it intends to answer that pressure. Explore 2026 laid out its response across three fronts: trust, discovery, and distribution.
Confirming "Plan with AI, Book with a Trusted Brand" in Numbers
The foundation of the strategy was not a flashy demo but a piece of research. Ahead of the conference, on April 14, 2026, Expedia published findings it titled the AI Trust Gap, based on a YouGov survey of more than 5,700 adults across the U.S., U.K., and India conducted in March 2026.
Acceptance of AI at the planning stage is advancing steadily. 53 percent are comfortable with AI suggesting travel options, 42 percent would use it to monitor prices, and 40 percent would use it to build itineraries. Booking is a different landscape entirely. 68 percent said they would choose a trusted travel brand over an AI chatbot to book, 66 percent said they would not trust AI to book on their behalf at all, and only 8 percent said they are comfortable booking through AI.
The anxieties are measured in concrete terms as well. Concern about losing control came in at 57 percent, payment and data privacy at 57 percent, misuse of personal information at 56 percent, and weak support when something goes wrong at 40 percent. What is being questioned is not the accuracy of the technology but the terms of delegation.
Travelers don't have a technology problem with AI. They have a trust problem.
CEO Ariane Gorin made the same point in the company's wrap-up for partners: this moment is not just about becoming more intelligent, it is about becoming more trusted. What the numbers show is that discovery and checkout are migrating to AI at entirely different speeds. The split between discovery and checkout is a structure shared across AI commerce, and the gap deepens for expensive, hard-to-reverse purchases like travel. Put the other way, brands that can shoulder accountability for payment and post-purchase support keep a clear role in the AI era. Expedia wove that gap into its strategy not as a threat but as its own position.
Content Becomes Product Data That AI Reads
The centerpiece of the partner-facing announcements was a set of three AI agents being built into Partner Central, the partner dashboard, introduced in the keynote by Chief Product Officer Shilpa Ranganathan. Partner Companion detects performance issues and commercial opportunities, Content Agent finds and fills gaps in listing information, and Autonomous Distribution speeds up onboarding by pre-filling listings from trusted sources. A single design philosophy runs through them: operational bottlenecks get cleared pre-emptively by agents rather than by manual labor.
From an AI commerce standpoint, Content Agent is the one that matters most. Photos, amenities, neighborhood detail, and policies have until now been material that supports a human booking decision. From here on, they are also the material an AI uses to decide which options deserve to be recommended. A product whose content is unstructured and incomplete loses human confidence and AI discoverability at the same time.
Expedia gave this shift a name: B2A, business-to-agent. Where traditional SEO optimized for human search, the question now is whether your inventory appears on the AI surfaces where planning and purchasing actually happen. The company also announced a collaboration with OpenAI Ads, Trip Matching with Meta, and an expanded Travel Media Network, moving to secure discoverability on AI surfaces through both advertising and distribution.
Translated to physical e-commerce, this maps directly onto product data readiness. Are product names, attributes, inventory, prices, and reviews organized in machine-readable form? That quality decides whether you make it into an AI's recommendation set. Content is no longer one element of marketing but the product data that carries AI-driven revenue.
Moving to the Supply Side of External AI Channels
The other pillar is turning the B2B business into infrastructure. According to Expedia Group B2B's announcement, the company already serves 75,000 partners and 200,000 travel advisors, and its platform processes 21 billion API calls a day. On top of that base, it previewed an AI toolkit still in development, rolling out with select partners in the coming months, designed to make it easier to connect Expedia Group capabilities into partners' AI experiences across APIs, interfaces, and agent workflows. At its core, the Intelligent Experience Platform is described as a set of composable AI components for assembling branded travel experiences.
Alfonso Paredes, President of B2B, said that with one connection, partners can build more complete travel experiences with less complexity. The target is the expansion of embedded travel, in which banks, retailers, loyalty programs, and fintechs that are not in the travel business embed travel into their own customer experiences. The agreement to acquire CarTrawler, an Ireland-based ground-mobility platform, expected to close in the second half of 2026, and the preceding Tiqets acquisition, announced in December 2025, are groundwork for extending the lodging-centered Rapid API to cars, flights, activities, and trip protection.
There is movement around MCP (Model Context Protocol), the specification for connecting AI to external systems, but its current state needs to be read precisely. What the conference showed was a developer-facing Dev MCP server that supports documentation, integration testing, and troubleshooting, alongside a B2B Labs showcase of MCP and other infrastructure for embedding Expedia functionality into partners' AI assistants. There is no announcement that external AI agents can autonomously book and pay for Expedia inventory over MCP. The toolkit itself is still in development, and how far support for connectivity standards will extend remains a question to verify. For the broader landscape, see our overview of MCP, A2A, and UCP.
The direction, even so, is clear. If the entry point of discovery disperses from owned sites to AI, move to the supply side so your inventory flows no matter which entrance wins. On its own apps, Expedia keeps polishing the direct experience with traveler-facing AI such as Virtual Agent for disruptions and AI Compare for weighing options, while opening inventory and capabilities outward. It is running the walled garden and the open garden on both fronts at once, from the marketplace's seat.
What Retail and E-Commerce Operators Should Take Away
Bringing Expedia's posture back to your own business, the questions narrow to three.
First comes meeting the conditions for being discovered by AI. As the term B2A signals, the quality and structure of product content is no longer a matter of presentation for humans but of whether you can ride AI-mediated channels at all. Missing attributes and inconsistent naming translate directly into dropping out of recommendation sets. The same groundwork also underpins any AI-assisted selling you build on your own channel.
Next is designing the journey around the trust gap. Even as users hand discovery and comparison to AI, they remain cautious about delegating payment. Rather than rushing a fully autonomous purchase experience, the practical near-term structure is to be discovered by AI while purchase and post-purchase support are handled dependably on a trusted channel. Showing who takes responsibility when something goes wrong is what moves conversion.
Last is re-examining your relationships with platforms. Marketplaces are reshaping themselves into the supply infrastructure of the agent era. Through which platforms does your product reach AI channels, and on what terms for fees, customer relationships, and data ownership? The further this infrastructure shift progresses, the more decisive those terms become.
Conclusion
What Expedia showed at Explore 2026 was not a defensive crouch against AI but a re-staking of its position as an intermediary. It confirmed with trust-gap numbers that booking still belongs to brands, made machine-readable content a revenue condition for its partners, and pressed ahead with infrastructure that carries inventory and capabilities into external AI channels. It stands as a case study in the moves available to a marketplace at a moment when discovery and checkout are shifting asymmetrically. For retail and e-commerce operators, the takeaway is to hold on to both product data that AI can read and customer touchpoints that carry trust. When the final step of booking or purchase begins moving to AI, the question is whether you can absorb that shift on your own terms.
