No role for legacy systems in AI future

The next major platform shift - one that occurs roughly every 15 years in the technology world - is already here, and it poses an existential threat to existing systems and investment strategies in corporate travel. That was the candid assessment delivered by Johnny Thorsen, VP of Strategic Business Development for Serko, at the FACTS conference, in a presentation that focused sharply on the travel buyer’s distinct and often conflicting agenda.

Thorsen’s core message was that the new AI landscape is so fundamentally different from previous shifts that its development is currently "uncontrolled" and set to move at a relentless pace, raising questions about the longevity of existing solutions, including online booking tools and distribution standards like NDC.

The Unique Power of Generative AI

The shift is unlike anything seen before because this new technology possesses the ability to create independently. As Thorsen noted, while past technological advancements relied on physical devices or mobile hardware, AI can "rebuild more by itself".

This autonomous development capability, coupled with a complete lack of current regulatory oversight, is accelerating the speed of change beyond all forecasts.

"Politicians are not equipped, knowledge-wise, to put a new legislation framework in place here," Thorsen stated, warning that the industry is likely to see "an uncontrolled development over the next few years".

Evidence of this acceleration is already in the numbers. Investment in AI data centres and infrastructure for the year exceeded the aggressive 2025 forecast by almost 40% by September, meaning "the speed of change will increase". This rapid, self-directed development is the engine that threatens to bypass current complex distribution layers.

For technology providers (TMCs and suppliers), this environment creates a precarious investment dilemma. Historically, early leaders often "disappeared" during platform shifts. The sheer pace of innovation means companies risk burning capital if they back the wrong horse, he warned.

This period of uncertainty means that while the technology is moving quickly, many established players may initially hold off on deep, risky investments, waiting for a clear winner to emerge.

The Leapfrog Protocol: MCP

The mechanism that could enable AI to leapfrog existing systems is the new global technology standard known as MCP (Model Context Protocol), an open source standard for connecting AI applications to external systems.

“Consider MCP the equivalent of the USB-C port on your laptop – it creates a global standard for accessing and using data in the Agentic AI world, reducing development time and cost,” he said.

Thorsen said MCP allows AI agents — the new robotic software platforms — to communicate directly with existing systems and data. This standard is important “because it means we are reducing the cost of innovation dramatically by having a standard early on”.

By providing a universal language for AI to interact with airline, hotel and booking data, MCP could potentially allow AI-driven agents to construct and manage trips with minimal reliance on legacy booking tool interfaces or the intricate, industry-specific standards (like NDC) currently being rolled out. In essence, a new, simplified data pipe is being established for AI, effectively allowing it to circumvent and render unnecessary the complexity of the old pipes.

Real-World Disruption: TravelX

One powerful example of new infrastructure providing incredible capabilities is the new airline ticketing system called TravelX.

This Spanish-based technology is currently being used by six low-cost airlines and is built on blockchain (specifically, NFT technology). This decentralized architecture enables flexibility impossible in the legacy world:

·       Ticket Flexibility: You can make a booking without putting the final passenger name or dates on the ticket initially. It's a seat token that can be named later.

·       Resale and Exchange: Travelers can resell the ticket if they can't travel, allowing them to recoup their money (with the airline taking a cut).

·       Pure Retailing: The airline can dynamically reach out to the traveller and offer a refund plus a seat on an alternate flight if demand is very high for the original ticket.

This model of pure retailing completely eliminates traditional legacy processes like cancelation, exchange, refund, and ticket settlement (BSP/ARC). This new freedom for the airline does not fit into the NDC standard in any way, underscoring the point that new solutions are already outpacing the latest industry standards. While it's a "low-cost game right now," several main airlines are reportedly reviewing the technology due to its benefits.

By providing a universal language for AI to interact with airline, hotel, and booking data, MCP could potentially allow AI-driven agents to construct and manage trips with minimal reliance on legacy booking tool interfaces or the intricate, industry-specific standards (like NDC) currently being rolled out. In essence, a new, simplified data pipe is being established for AI, effectively allowing it to circumvent and render unnecessary the complexity of the old pipes.

Thorsen’s most urgent message for the travel industry is that AI leapfrog is imminent. The speed of change will accelerate drastically, threatening to render existing distribution standards like NDC, and the booking tools built upon them, obsolete. This rapidly evolving environment creates a "Make-or-Break" Technology Investment dilemma for providers like TMCs and suppliers. Historically, leaders who bet on the wrong platform during major shifts often "disappeared," meaning current investment carries a high risk of burning significant capital and time.

Thorsen left the audience with plenty to think about in his predictions for the corporate travel environment for 2030:

·      TMC revenue will be reduced significantly, but the profit margin will increase even more

·      The 2030 TMC will be a TDC (Travel Data Company) with a range of digital services

·      The human agent pool will be reduced by 50% or more

·      The supplier community will all have MCP servers enabling access for Agentic AI solutions

·      Re-shopping bots will consume public pricing non-stop, motivating airlinesto move to smart tickets

·      Payments will shift to blockchain-based stablecoin transfers with credit card payment becoming a back-up

·      Human-driven search will be replaced by trip triggers which activate your preferred travel bot

·      Leisure travel will lead the way as there are less complications in terms of policy, cost control and infosec

·      Web-based travel services will be a legacy model as we go "bot on mobile" first.

Despite all of the tech advancements, people will remain vital in corporate travel, Thorsen concluded, but human expertise must transition to strategic, relationship-focused, and complex exception-handling tasks that AI cannot yet master.