Matthew Newton is chief architect at travel management company CWT
Following on from predictive AI and generative AI (Gen-AI), a “third wave” of artificial intelligence is now emerging that will fundamentally change how we work: Agentic AI.
Unlike Gen-AI, which requires human prompts to generate content, Agentic AI can – among a wide range of uses – autonomously manage travel bookings, optimise itineraries, and handle disruptions in real time. It can set its own goals, make decisions, and acts without human intervention.
While Gen-AI focuses on creating content, Agentic AI is designed to form judgements and can collaborate with other AI agents to achieve complex tasks. Its use cases in our industry are diverse and its potential impact on travel management is considerable.
Consider a future where your personal AI agent contacts the AI agent of a car rental company to negotiate the best deal based on your preferences and schedule. Does that mean the end of human travel managers? To borrow a phrase from Mark Twain, “reports of their demise are greatly exaggerated.” While AI will undoubtedly take responsibility for a range of key tasks in our industry – just as it will in many others – this shift doesn’t mean the end of travel management as we know it. If anything, it opens the door to add even more strategic value to organisations.
GOVERNANCE ON AUTOPILOT
Booking compliance has always been a balancing act between policy enforcement and traveller convenience. Today’s leading travel platforms – like Egencia, myCWT, Navan, Spotnana and many others – already do a strong job of guiding employees toward in-policy choices. These self-service tools use rules-based filtering, approval workflows, and content configuration to shape compliant behaviour.
But Agentic AI takes things further, transforming compliance checks by introducing machine-driven governance. AI agents continuously scan and monitor bookings in real-time, flagging policy violations and automatically correcting them. It doesn’t just enforce policy after the traveller initiates a search – it can anticipate needs, orchestrates options, and adapts in real time. This is the shift from configured compliance to contextual intelligence.
With Agentic AI, travel no longer begins with a manual search; it starts with a prompt, a proposal, or even a fully formed itinerary. Preferences, schedules, policies, and even personal circumstances (like school holidays or wellness constraints) are factored in automatically. Travellers are guided – not governed – toward the optimal choice.
And when AI agents exist on both sides of the relationship – within the client organisation and the travel management company – booking compliance becomes a coordinated system of intelligence. These agents don’t just operate in isolation; they communicate machine-to-machine, aligning schedules, cross-checking policy boundaries, optimising inventory, and even adjusting itineraries on the fly when plans change.
This opens up new levels of efficiency. AI agents from the TMC can sync with client-side agents to ensure recommendations reflect both contractual obligations and evolving company goals. Meeting bots can coordinate across calendars before suggesting travel options. Service agents can negotiate with suppliers and respond to disruption dynamically – no human intervention required.
But this isn’t about replacing the human layer – quite the opposite. Agentic AI frees up travel managers and TMCs to focus on the strategic levers: policy evolution, supplier strategy, traveller wellbeing, sustainability goals. It’s a force multiplier, taking care of the transactional and contextual so that people can focus on the exceptional.
THE FUTURE OF TMCs
If Agentic AI can autonomously book travel, align with policy, and coordinate with suppliers, then a reasonable question arises: why would an organisation still need a travel management company at all? It’s a fair challenge but it misunderstands both the nature of Agentic AI and the purpose of managed travel.
Agentic AI is not about deploying technology for its own sake; it’s about solving real business problems. And like travel programmes themselves, the way AI is applied will vary significantly from one organisation to another. Some may invest heavily in developing their own AI agents. Others will rely more on external partners. But in every case, the same truth holds: building, training, and maintaining intelligent agents that deliver a high-quality, trusted travel experience is not trivial – and it is certainly not free.
Travel sits outside the core business of most companies. Developing and operating domain-specific AI services that require nuanced supplier relationships, deep inventory access, contextual policy enforcement, and 24/7 responsiveness is not just a heavy lift – it risks creating new complexity rather than solving for it.
A trusted TMC, equipped with its own Agentic AI capabilities, brings a series of immediate benefits. It delivers scale and specialisation across travel content, booking workflows, disruption management, and duty of care. Continuously evolving AI agents can stay aligned with fast-changing models, user behaviors, and supplier dynamics. And human-in-the-loop oversight can ensure bias, hallucinations and opaque decision-making don’t derail the user experience or compromise corporate goals.
Agentic AI changes how TMCs deliver value but it doesn’t erase their relevance. In fact, it deepens it. Because in a world of intelligent agents, the companies who understand which partnerships to optimise – not just which models to train – will be the ones who move fastest, scale smartest, and deliver the most resilient travel programmes.
THE ROLE OF THE TRAVEL MANAGER
If Agentic AI can orchestrate policy-compliant bookings and manage real-time interactions between organisations and travel providers, what’s left for the travel manager to do? In short, everything that matters most.
Agentic AI doesn’t eliminate the need for a travel manager – it elevates and redefines the role. Just as a TMC becomes a strategic partner in an AI-enabled ecosystem, so too does the internal travel manager remain essential to the governance, design and evolution of the travel programme itself.
Because Agentic AI, at its heart, is not just a technology service, it’s a new kind of relationship. One that requires clear intent, mutual commitment and continuous calibration. Someone must own the design of that relationship. Someone must ensure it aligns with business strategy, corporate values and evolving expectations across regions, user groups and lines of business. That person is the travel manager; someone shaping the blueprint.
With AI working on the client’s side, travel managers can shift from enforcing compliance to orchestrating value. They can define differentiated service models for distinct user populations – executives, road warriors, occasional travellers – and let intelligent agents coordinate seamlessly across multiple travel management providers, if necessary. Agentic AI becomes a toolkit for more sophisticated segmentation and smarter service delivery. But only if someone is shaping the blueprint.
Ultimately, the travel manager becomes the strategic architect of the travel programme, ensuring that Agentic AI is used intentionally, ethically and in ways that are aligned with the organisation’s long-term goals.
The future is not about Agentic AI replacing humans, but rather augmenting human capabilities to focus on higher-level strategic work and relationship building. The organisations that thrive in this AI-driven future will be the ones that successfully integrate AI into their culture and adopt economic models that make adoption feasible and scalable.