Summary: AI in corporate travel has moved from pilot projects to daily practice, and travel managers use it most for pricing, booking, and traveler support. It still gets facts wrong, raises data security questions, and falls short in complex disruptions. The programs getting the most from it start small, keep people reviewing the output, and pair new tools with human expertise and strong supplier relationships.
Eighty-two percent of US business travel managers say they'd consider adopting agentic AI right now, according to “State of AI in Corporate Travel 2026,” a February 2026 survey of 302 travel managers by travel technology company Serko. Agentic AI goes beyond just suggesting travel options, though. It acts on them, rebooking a flight, reconciling an expense, or adjusting an itinerary within your company’s policy.
Still, that number needs context. Everyone in the survey’s sample already uses AI at work, and 98% are employed at companies with more than $100 million in revenue. The survey also shows where large, AI-ready programs are headed, which isn't the same as where every agency or mid-size company stands today.
Even so, artificial intelligence in business travel has moved from trial projects into daily operations, and travelers now bring AI-built plans to the booking conversation. The practical question concerns which jobs you should hand to AI and which you should keep firmly in human hands.
AI is already built into much of the corporate travel technology that travel teams use every day. In Serko's 2026 report, 80% of travel managers use AI for pricing optimization, 78% for booking efficiency, 66% for traveler-facing tools like chatbots, 63% for safety alerts, 55% for itinerary creation, and 43% for travel data analysis.
What travel managers want from AI has shifted, too. Efficiency now outranks cost: 37% name efficiency and productivity as their primary goal, ahead of better data analysis (27%) and a better booking experience (24%).
Cost remains a key concern. In GBTA's January 2026 Business Travel Industry Outlook Poll, 74% of US travel buyers cited program cost savings and control as a concern. AI is how travel teams plan to handle more work with the same headcount.
Travelers are moving faster still. Phocuswright found that 56% of US travelers used AI for at least one trip in the past year, more than double the 2024 rate. Sure, most of that is leisure travel, but those same people also book business trips.
Pricing is where AI travel management has the longest track record. AI tools do repetitive work faster than any person could, including: comparing rates across hotels and fares, catching out-of-policy bookings before they're ticketed, and spotting spend leaking away from preferred suppliers.
Travel managers are increasingly comfortable letting AI act on its own here. Serko found that 88% trust agentic AI to make flight changes, and 92% feel confident using it to generate and reconcile expense reports. For a traveler whose connection cancels at 11 PM, an automatic rebooking that arrives before they reach the gate is a genuine service improvement. One they won’t soon forget.
This is the largest untapped opportunity. Only 43% of travel managers in Serko's survey use AI for travel data analysis, while 27% name better reporting and insight as their main reason for adopting it. AI can work through months of booking data to show where spend is heading, identify which suppliers travelers use, and reveal where negotiated rates are going unbooked.
Some of the fastest wins for agencies happen away from the traveler: a first-pass client reply, a condensed supplier update, itinerary notes, and next steps from a meeting. In a small agency where one person covers several roles, those minutes add up rather quickly. Personalization is improving, too, with agentic tools recommending options from a traveler's history, and it works best when an advisor checks the result against what the trip requires.
AI can be confidently wrong, as even the casual user of common AI platforms has learned. In an August 2026 survey commissioned by a travel concierge company Greetwell, 1 in 6 AI-using leisure travelers had been recommended a place, tour, or activity that didn't actually exist. As a result, a client may arrive with an AI-built itinerary that's half useful and half out of date. Someone has to catch the difference, and it ought to be a human.
Travel data is sensitive: passport details, accessibility needs, booking references, and negotiated supplier rates, to name a few. Yet 84% of travel managers in Serko's survey use generic AI tools such as ChatGPT, Claude, or Gemini, compared with 39% using AI tools built for travel management. A consumer chatbot and an enterprise tool with security controls carry very different risks. Programs with travelers in Europe also need to account for GDPR and the EU AI Act.
Chatbots are the most widely deployed traveler-facing AI tool, and they aren't the best-rated. In BTN's 2026 Corporate Travel AI Survey, travel managers who had deployed chatbots rated their effectiveness 3.52 out of 5, below nearly every other tool in the survey. Again, automation handles routine requests quite well. However, it struggles with the traveler stranded by a storm that grounded every connection, with airport hotels sold out and a client meeting that can't be moved.
That call needs judgment, supplier relationships, and someone who picks up the phone.
Deciding to use AI is easier than making it work. 60% of travel managers in Serko's survey named technology and implementation challenges as their biggest barrier to adoption.
Nothing in the current data suggests it will. In fact, looking five years out, only 27% of respondents in GBTA's January 2026 poll expect AI to significantly transform their responsibilities, while 56% expect moderate improvements from automation.
Instead, the job is changing shape, with less time spent on routine tasks and more on interpreting information, asking the right questions, providing emotional support to clients and team members, and hopefully spotting risks earlier. A traveler rebooked twice in one night doesn't need another automated notification. They need someone who owns the problem, understands their frustrations, and gets people exactly where they want to go.
When a client shows up with an AI-generated hotel comparison, the advisor's value is knowing which parts of it are wrong and what that means for the human being experiencing it on the other end.
Start narrow. We recommend picking one high-value workflow, choosing enterprise-grade tools, and keeping a human reviewing results while you test. Time the task before you start, so you know what the tool saves.
A sensible sequence moves from internal productivity, like summarizing notes and drafting emails, to lower-risk traveler-facing uses such as general FAQs. Automated booking and servicing come last, after proper testing. Move to the next stage only when the current one runs reliably with light review.
The payoff is time, best spent on the work that still needs a person: advising clients and managing supplier relationships.
Hickory Global Partners' position is that AI earns its place when it gives travel professionals more time for the work that depends on people.
That people-dependent work is what Hickory has spent more than four decades building. Hickory Global Partners connects 1,900+ travel management companies (TMCs) and corporate travel departments with negotiated rates at 50,000+ hotels in more than 8,000 cities and 180 countries.
When a disruption needs a person instead of a prompt, Hickory Solutions365, Hickory's 24/7/365 travel fulfillment center, puts live travel advisors on the line.
On the technology side, the Hickory Marketplace introduces members to pre-vetted travel technology providers. That way, agencies and travel departments can evaluate new tools without having to vet every vendor from scratch.
Become a Hickory member today.