Generative AI is evolving the travel discovery journey from static search (research) to dynamic conversation (collaboration). Travelers no longer just browse for inspiration, they collaborate with generative AI systems to co-create their ideal itineraries. But that collaboration has limits. AI can recommend, refine, and organize. But it can’t execute...yet.
Part of that limitation is technological (more on that later). But another significant aspect is limited trust. Research from Bain shows that only 10% of consumers have purchased something through AI, and just 24% feel comfortable letting it complete a transaction. In travel, where purchases are high-value and personal, that caution likely runs deeper.
For AI to fulfill its role as a digital concierge, we will need to trust it. And AI will earn our trust by proving reliability, starting with small, repeatable successes.
As a research assistant, AI helps travelers explore ideas that fit their stated preferences. This is where we started in the early days of generative AI tools (ChatGPT, Perplexity, etc.). Today, many consumers see AI as a collaborator that helps refine ideas and define viable options through conversation, narrowing thousands of possibilities down to a few strong choices.
From there, the next step is to authorize AI to act for us, to be our agent, booking trips for us and spending our money. But this won’t happen overnight. Only once AI has demonstrated a high level of accuracy and earned user trust will travelers feel comfortable handing over the proverbial keys.
The technological maturity of AI in travel will be a factor, too. Reliability, the kind that earns trust, can only be built on something travel still struggles with: a strong, efficient orchestration layer. For AI to be dependable, it needs clean, connected data; a unified foundation that links intent, identity, and execution.
The Quiet Revolution of Orchestration
Orchestration rarely gets the spotlight, but it has powered nearly every major technological leap for decades. Sitting between user-facing applications and back-end infrastructure, an orchestration layer translates data between systems, reconciling differences in how they store or define that data, and triggering the right actions in real time. Across industries, it has repeatedly served as the bridge between intent, identity, and execution, unifying disconnected systems that hold back innovation.
In financial services, that innovative shift was open banking. For decades, banks operated in closed systems that controlled customer data and the tools built around it. Users couldn’t easily move money, compare rates, or use digital services from institutions where they didn’t have accounts. When the European Union introduced the Payment Services Directive (PSD2) to break down these barriers, banks had to find ways to let customers share data securely.
An orchestration layer translated data between the legacy infrastructure of traditional banks and the modern APIs that fintech apps use. Companies like Plaid and Tink emerged, connecting financial institutions to digital tools that could act instantly on verified information.
The results have been transformative. More than 90% of financial institutions can now use APIs and 61% say real-time payments are key modernization drivers with benefits like 30% lower IT spend, 10% higher client acquisition, and 25% lower operations costs. Transactions that once took days now happen instantly. And consumers never see the complexity. They experience reliability, and trust follows.
Travel’s Fragmented Foundation
Travel is poised for a similar breakthrough. The industry’s building blocks, like loyalty programs, supplier feeds, booking engines, and payment systems, each perform well individually, but rarely in unison. Upgrades are missed because systems don’t sync, dining suggestions are detached from itineraries, and loyalty points are left unused. Each disconnect weakens trust because travelers don’t experience a continuous thread between what they want, who they are, and what the system can deliver.
The iSeatz Modern Traveler 2025 report found that 31% of travelers cite a fragmented booking process as a top frustration, ranking just behind hidden fees as the most common pain point. This reflects a widespread breakdown across multiple systems that travelers encounter at nearly every step of the journey. Even one disconnected hand-off erodes trust in the entire experience.
Our research also found 73% of travelers are more likely to book with brands that use AI for personalized recommendations. That expectation puts pressure on the entire ecosystem: AI can understand what travelers want, but it can’t deliver on true personalization unless the underlying systems are easily accessible and able to produce coherent and consistent recommendations across a fully connected trip.
An effective orchestration layer makes that possible. It can normalize data across sources and ensure consistent delivery to every connected system. But in the absence of standardized APIs that normalize supplier data, apply consistent business logic, and ensure real-time synchronization across booking, loyalty, and payment platforms, today’s orchestration relies on one-off APIs to connect to disparate systems.
With the advent of AI and the promise of MCP (Model Context Protocol), there is the potential for much simpler and sustainable connections. MCP creates a universal protocol for AI systems that adopt it to safely and dynamically access external data sources like loyalty databases, supplier APIs, or booking engines. This could mean an AI assistant that doesn’t just reference a static dataset but can pull live room availability directly from a hotel’s reservation system, check loyalty point balances, and feed that information into a unified itinerary while preserving data privacy. Instead of dozens of custom integrations between systems, MCP could enable a shared standard for how AI “talks” to the travel ecosystem. If orchestration is the foundation, protocols like MCP could be the accelerant, reducing technical friction, improving reliability, and enabling the next generation of AI-powered travel experiences to scale sustainably.
Imagine the difference. Today a Gen AI tool might say, “Here are three hotels you might like.” Tomorrow it will say, “Here are three hotels that match your loyalty status, reduce transfer time, and can be booked with points. Which one would you like me to reserve?” That’s not full autonomy; it’s collaboration. And that’s a crucial step on the way to a true concierge experience.
From Interaction to Confidence
The prerequisite to trust isn’t blind faith in automation; it’s proof of reliability. Travelers will fully trust AI when it surfaces the right answers enough times to feel dependable. The more consistently it gets micro-interactions right by bringing intent and identity together effectively, the more willing travelers will be to let it handle execution.
The path from researcher to collaborator to concierge isn’t a dramatic leap. But authorizing AI to act for us and be our agent is a climb that can only be built on reliability, reinforced by trust, and scaled by smarter orchestration.
Want more news and insights from iSeatz? Sign up for our newsletter!
To learn more about our work in travel technology, contact us here.