When a winter storm shutters a major hub or an airline IT failure ripples across continents, the cost of disruption goes far beyond delays, spilling into missed meetings, stranded travelers and overwhelmed corporate travel teams scrambling to rebook en masse. For decades, disruption management has been a reactive exercise, reliant on fragmented data and manual rebooking. Today, artificial intelligence, especially the growing use of generative and agentic AI tools, is changing the game. 

From predicting irregular operations before they cascade, to automating booking changes in real time, AI has the potential to shift disruption response from triage to strategy. But as travel managers and travel management companies explore these capabilities, a critical question emerges: can AI deliver not just speed, but control when it matters most? And are corporates ready to trust these tools? 

BTN’s survey indicates at least some appetite for utilizing AI for travel disruption management, with 15 percent of respondents stating they have implemented AI within their travel programs to address this issue. 

Clinical research organization Parexel together with its TMC partner in March launched a pilot program to implement AI-powered monitoring to better predict the risk of travel disruptions. 

"We've observed a clear increase in travel disruptions, from winter storms and strikes to geopolitical dynamics. Our approach needed to shift from reactive to proactive," explained Parexel executive director of travel and sustainability Ben Park. 

The pilot currently involves 90 frequent travelers, including senior leaders and sales personnel. AI agents monitor flight bookings and assess disruption likelihood based on flight patterns, major weather events and historic airport data.  

When the system identifies potential disruption, AI agents alert Parexel's TMC partner and monitor the traveler's journey with greater scrutiny.  

For high-risk scenarios, Parexel’s executive assistants and the travel management team are also notified via email. A dashboard also displays at-risk travelers using a traffic light system for quick visibility. Any necessary booking changes, however, are made by the TMC’s human travel agents. 

"The AI system monitors patterns and alerts our team, but human travel agents retain decision-making authority. AI augments human expertise—it doesn't replace it," Park said. 

"We're in active testing and discovery phases. Currently, the system identifies disruptions as they occur. Our next phase focuses on predictive capabilities to anticipate disruptions before they impact travel," Park explained.

A U.S.-based talent agency has been utilizing AI to assess potential travel disruption via its TMC partner since 2024. 

The agency’s head of travel, who spoke to BTN on the basis of anonymity, said pre-trip alerts and on-trip notifications are sent directly to travelers via a dedicated mobile app, but the solution does not handle booking changes. The travel manager noted, however, that travelers have to opt in to receive notifications and are required to download a separate app on their mobile device, which has hindered widespread adoption. 

"It's a separate component that needs to be downloaded [on travelers' phones],” they said. "I think that's what the issue has been, because travelers have their airline apps, the TMC app... it's too much.” 

This travel manager said a human-in-the-loop process will remain key for their program and that most servicing will continue to be handled by a human travel agent. 

 "We are more cautious, and we're more of a white-glove operation,” the travel manager said. The company’s multigenerational workforce also means that “we’re not one-size-fits all… some [travelers] still prefer to speak to someone and want their boarding passes printed, while others prefer to do it themselves,” they said. 

Elevance Health, meanwhile, works with flight data technology provider Lumo to help predict potential travel disruption for their VIP travelers in the U.S. as well as for large meetings and events. Lumo uses AI to score flights on a one-to-10 risk index, considering factors like recent flight history, seasonal impacts, changing schedules and flight patterns. 

From a tooling standpoint, executive assistants and meeting planners are provided with a dashboard that monitors travelers’ flight schedules. They also receive alerts when a potential disruption is assessed at a risk factor level of eight. 

“Any time we have a large group of people meeting in a city, we like to have that ability to predict and be proactive with any potential flight disruptions… because we found that for some meetings, we would be behind the ball with a lot of people pushing us for changes all at once, and that was unmanageable,” said Elevance travel manager Travis Steed.  

And in the case of VIP travelers, Steed says “it's nice to have a back-up plan.” 

Steed also stressed the importance of Elevance owning its relationship with a third-party supplier like Lumo, rather than relying on its TMC for AI innovation. 

“As a general rule, we like to have those direct relationships with vendors versus being reliant on the TMC and the objectives they have with those vendors,” Steed says.

Elevance has also updated its contracting language to avoid “any kind of block or hindrance coming from the TMC” regarding sharing real-time booking data with third-party vendors.  

“We consider booking data to be our data. So, if we choose to work with a vendor, we expect whoever that [TMC] partner is to send that booking data so that we can have these independent relationships regardless of what the TMC has participating in their vendor supply chain.”

Trust Issues 

In all the use cases above, AI is providing assistance but is still not being used to trigger a booking transaction when a disruption occurs, despite the technology being available. 

“It’s not a technology readiness problem. The capability largely exists today,” said Martijn van der Voort, director of consultancy AstraNomad and former CWT director of product delivery technology. “I think the real roadblock here is the organizational willingness to redesign [legacy systems] because that means directly challenging the call center economics of a TMC, and the control structures that certain leaderships are not prepared to let go of.” 

“The vast majority of disruption is a simple rebook or cancellation. It's a policy-governed decision with a very limited set of outcomes. We don't need humans for that,” he said. Still, he differentiates between ‘routine’ disruptions and more complex crisis scenarios.  

“AI is not going to physically extract people from a conflict zone or a collapsing security situation–but then again, neither will your TMC. … That is a travel risk and a security function. It's not a booking function.” 

A further challenge to widespread AI adoption in corporate travel is the uncertainty surrounding what the technology can actually do. 

“There are so many false prophets out there that are that are claiming tools are agentic AI, and they're frivolously and liberally using several AI terms and mixing them all up, which means the confused [travel managers] remain confused,” van der Voort said. “That’s very dangerous because, what do people do when they're confused? They fall back on their old ways of doing things.” 

Park agrees that more education is needed to ensure travel managers are properly informed around AI’s capabilities. 

"There's significant enthusiasm around AI right now. Our focus is demonstrating measurable, practical value. We're committed to clear communication about what's achievable today versus future capabilities," Park said.