Blogs
6 mins read 
At this point, one thing is undeniable: everyone in the business travel ecosystem is using AI in some way, shape, or form. Some are experimenting quietly. Others are pushing aggressively. And a growing number are already building real workflows around AI to reduce manual work, improve traveler experience, and strengthen program oversight.
But if the possibilities are endless — and they are — then the bigger question becomes: what actually scales in managed travel?
That’s exactly what surfaced in our latest Business Travel Innovation Group (BIG) session. The conversation wasn’t about “AI for AI’s sake.” It was grounded in practical realities: where AI is showing value today, where trust breaks down, what data corporates truly need, and how modern airline retailing may serve as an enabling layer for what comes next..
AI Is Here — But It’s Not One Single “AI Strategy”
The strongest takeaway from the session was that there isn’t one AI story — there are many. Buyers, TMCs, airlines, and tech providers are each approaching AI through the lens of their own operational constraints, risk tolerance, and desired outcomes.
Some participants are focusing on immediate productivity wins: using AI to support tasks that are traditionally human-heavy and tedious, like price spot checks, benchmarking, contract review, or pre-trip approval logic. Others are exploring how AI agents might help manage complex itineraries, auditing workflows, and approvals — with many stressing the importance of keeping humans “in the loop,” especially for compliance-related decisions.
Across the discussion, a pragmatic consensus emerged: AI can accelerate work, but it doesn’t eliminate accountability.
“Trust but Verify” Is the New Operating Model
One phrase captured the tone perfectly: trust but verify.
Several participants noted that AI can be extremely helpful — but also inconsistent, especially when it comes to math, validation, and precision. Many shared the same approach: AI is often easier to tweak than starting from scratch, but outputs still need to be checked.
This dynamic matters even more in corporate travel, where decisions directly impact spend, duty of care, compliance, and negotiated program value, especially as airlines transition toward offer-order transformation (OOSD). If AI is going to scale in managed travel, it will require a model that supports verification — not blind automation.
NDC, Modern Offers, and the “Why” Behind Savings
When the discussion moved to NDC and value realization within modern airline retailing, participants challenged assumptions in the healthiest way: What does savings actually mean, and compared to what?
The group discussed industry-cited savings ranges (often referenced in corporate contexts) and clarified that these are typically benchmarked against legacy content flows (e.g., EDIFACT/GDS). What’s driving those differences isn’t magic — it’s structural aligned with frameworks such as IATA’s NDC and order-based retailing vision: continuous pricing, dynamically created offers, bundled options, and cases where some carriers have removed certain low fares from traditional channels.
One important point: while hard-dollar savings are attractive, many noted that experience and servicing outcomes may be the real differentiator — especially as offers become more customized and order-based.
Direct vs Indirect May Matter Less — Until Servicing and Data Are on the Line
A compelling debate surfaced around whether it will matter in the future if NDC is accessed directly or via intermediated channels, a question closely tied to how ecosystems like Connected Travel Sellers enable access to modern offers at scale. If AI becomes the interface — the layer that searches, compares, and recommends — does the “pipe” matter?
The answer from the group wasn’t binary. Some felt the distinction matters less as long as the system can access and execute modern offers. Others emphasized that servicing and change management remain a key difference today, with direct connect NDC sometimes proving easier for modifications and resolution.
In other words: distribution models may blur — but execution quality will still separate what works from what doesn’t.
The Real Battleground: Data, Privacy, and Who Gets Visibility
If trust is the operating model, then data governance is the foundation.
One of the most thoughtful moments in the session centered on order and event data: who owns it, who stores it, and who can access it — corporates, TMCs, airlines, or platforms. Participants explored what’s technically possible in modern, cloud-based order management environments, while recognizing that access can still be constrained depending on airline permissions and ecosystem roles.
Just as important: several participants emphasized the need for tokenization and anonymization, especially as AI introduces new “context” into the shopping and servicing journey. If AI captures how a traveler or arranger asked for something in natural language, does that create new forms of behavioral data that could influence offers?
The group agreed that protecting privacy must be designed in from the start — not bolted on later.
MCP, Agents, and a Future Where Corporates Build Their Own Workflows
The conversation also touched on MCP (Model Context Protocol) and the growing concept of “AI agents” — systems that can maintain context, interact with APIs, and orchestrate tasks without heavy custom development.
What stood out wasn’t the acronym — it was the implication: corporates could increasingly customize AI workflows aligned to their own travel philosophy, rather than adopting one-size-fits-all dashboards and processes.
That possibility is powerful — but it reinforces the same theme: if corporates build more capability, the ecosystem must also evolve the frameworks for trust, auditability, and verification.
The BIG Takeaway: We Learn Faster When We Learn Together
This session reinforced why BIG exists. No single stakeholder has the whole picture (a reality increasingly reflected in broader industry analysis from outlets such as PhocusWire). Airlines see one set of constraints. TMCs and OBTs see another. Corporates live the operational reality every day. And technology is moving faster than policy and process can keep up.
But when the ecosystem shares openly — through discussion, chat, and follow-up conversations — we start assembling the puzzle more quickly.
AI is not a question of “if.” It’s a question of how: how we validate, how we protect trust, how we govern data, and how we ensure AI-driven travel outcomes remain practical for managed programs.
Because in business travel, what matters isn’t hype.
It’s what scales.
Press Releases
Whitepapers
