On the inevitable bots versus humans race, generative AI is a persuasive new tool when it comes to improving the customer experience.

Personalised travel planning and AI-driven concierge services are already here and the emergence of the Digital Assistant is not far off. However, advances are being tempered by the naturally risk-averse TMC industry.

“As a TMC, we’ll never lose sight of the importance of human connection. We see AI as an opportunity to help our people work smarter and faster, focusing on high-impact tasks that drive traveller (and client) satisfaction,” says David Coppens, SVP Solutions Engineering and Service Design, BCD Travel. “While we’re leveraging AI to support requests throughout various departments, our technologies will always have human oversight to monitor and measure precision.”

Human nature

AI can certainly reduce high volume, low-value tasks through automation, and high-touch TMCs have been working out how best to marry their human workforce with artificial intelligence. “While human agents and AI agents will always overlap in some capacity, there will still be a need for human agent expertise,” reckons Navan’s CEO Rich Liu.

AI will maximise the productivity of ‘real’ agents and their jobs will change as a result. As repetitive tasks are automated, agents can focus more on more complex issues that work best with human interaction, such as crisis management, VIP services and complex travel bookings.

Moreover, how much AI is utilised will be influenced by company culture, travel policy and the travel manager.

This ‘slowly slowly’ approach with AI is not surprising as TMCs work in a space where accuracy is of the utmost importance and nuance is everything. TMCs are following the low risk/high impact productivity route first while they test and learn.

”We deal with payment, company travel policies, personal preferences, PCI/PII data and so on,” says Coppens. “We are focused on fairness, transparency, privacy, security, responsible use and human oversight. It is what our clients expect and our AI use-cases are carefully selected to ensure we meet that expectation.”

“It is slow but it’s the right choice as there are dangers of generative AI and TMCs have to protect themselves,” adds Nonna Alomaia, Chair of the Technology Working Party at the Business Travel Association (BTA). “Mishandling data is the issue.”

Thinking ahead

Fast forward past the test phase, though, and AI has near-endless potential to change the way our industry processes travel bookings, from search to expense and reporting.

Once AI learns what the travellers’ preferences are, its predictive tools will anticipate needs and create a bespoke booking that is policy compliant before travellers even think about their next trip. AI has pattern recognition that will make light of any trip a traveller has taken before.

Sarah Kuberry Martino, Direct Travel’s Chief Product Officer, believes this hyper-personalised picture of the future is closer than ever.

“We envision AI going beyond simple voice commands. Not only will travellers be able to dictate preferences like, ‘I need to go to London Tuesday night and return Friday morning,’ but AI could generate suggested itineraries proactively based on calendar integrations, meeting locations, and historical data. This level of anticipatory service will help transform booking from a task into a largely automated, yet highly personalised, transaction.”

BCD’s Coppens agrees. Based on the current speed of development, he says: “Agentic-AI, where AI acts on behalf of the user (rather than just assisting the user), and autonomously makes bookings without user input, taking into account travel policies and preferences, does not seem like a distant future.”

The BTA’s Alomaia says the AI Predict tool of her previous employer, Altour, can check a number of data sets – from weather, airports and traffic, for example – and predict what the likelihood is of a flight delay as the cabin crew could be stuck on the M25. ”It’s the most impactful tool it has. We had a traveller advised of a delay three hours before the airline knew.”

CTM reckons its virtual travel assistant, Scout, could deliver more services in the future. “Features like recognising loyalty programmes, seat preferences, and past room numbers are part of the vision for enhancing traveller satisfaction and streamlining the booking process. All the traveller would need to do is approve Scout’s suggestions,” explains Eric Ediger, the TMC’s Global Head of Automation Machine Learning.

In the meantime, the easy wins are enhancing the virtual travel assistant. At Navan, the company is already averaging a 20% year-on-year increase in tasks completed by its automated agent, and by next year, “it will likely be managing 70% of support requests,” says CEO Rich Liu.

“This shift will enable travel agents to become even more specialised in their craft. We’re betting big on chat becoming the primary interaction mode, with increasingly sophisticated voice capabilities integrated into AI bots.”

Already, Navan’s automated support agent is able to anticipate users’ needs based on known traveller patterns and share traveller-specific recommendations.

“This level of personalisation takes the booking process from hours on a legacy platform to minutes or seconds with a modern T&E solution,” says Liu.

Price alterations

Clearly, these labour and cost savings will impact transaction fee and fee models. “Yes,” says Direct Travel’s Kuberry Martino. “As AI begins to anticipate needs and generate personalised, book-ready itineraries, we expect a significant shift in how travel services are priced.

“Traditional per-transaction fee models – built around manual booking processes – become less relevant when automation handles much of the work. At Direct Travel, we’re already exploring per-trip pricing, which reflects the value of the entire travel experience rather than individual tasks. This approach aligns better with AI-driven workflows while still recognising the importance of expert human support throughout the journey.”

ATPI Group Commercial Director Katie Skitterall concurs. “TMCs with AI-driven solutions will shift toward a single transaction fee that includes booking, unlimited changes and cancellations, which is very different to what is happening today,” she says. “However, this evolution will require careful management, particularly when it comes to buyer expectations. Comparing fee models will be more complex and many businesses still aren’t effectively tracking changes.”

BCD’s Coppens disagrees: “TMC fees are a compensation for the work we deliver and the investments we make. I don’t expect a direct impact from the use of AI in that area. Continuing use of human oversight may keep the status quo, at least initially.”

Shifting attitudes

Aside from possible changes to fees, traveller engagement is another upcoming challenge says ATPI’s Skitterall. “For AI to deliver a truly bespoke booking experience, we need travellers to update their profiles and preferences, and businesses to commit to integrating their systems effectively. The next generation isn’t afraid and will push for a streamlined end-to-end virtual experience as the workforce demographic change.”

One thing is for sure, change is coming and TMCs will pivot, yet again, to keep ahead of the game.