Used effectively, AI can be one of the most powerful tools available to travel and hospitality companies for solving strategic problems. But it is not a strategy. That distinction is becoming more important because the cost of AI is now a genuine business consideration, not an abstract technology concern.

AI investment involves more than software licenses or token costs for model usage. Successful adoption needs to be supported by data integration, cloud infrastructure, cybersecurity, system modernization, employee training, governance, compliance and ongoing maintenance. Costs can also increase when AI is deployed at scale across millions of customer interactions or operational decisions. A proof of concept can be an effective way to validate a use case, but calculating its potential business impact requires organizations to consider real-world performance, integration with legacy systems and the level of ongoing human oversight required.

The economics of the industry make that discipline particularly important. In June 2026, the International Air Transport Association (IATA) forecast a global airline net profit margin of just 2% for the year, with net profit per passenger expected to fall to $4.50. Hospitality faces similar cost pressure. CoStar and Tourism Economics’ June 2026 US hotel forecast projects RevPAR growth of 2.8%, but expects expenses to rise at a faster rate, resulting in a continued squeeze on profit margins. In industries operating under this kind of margin pressure, technology investment needs a clear connection to measurable business value.

Focus AI on the problems that matter

This makes selective adoption essential. Not every customer journey needs an AI solution and not every process generates enough value to justify one. The strongest use cases are often high-volume, data-rich and operationally measurable.

Airlines, for example, could prioritize predictive maintenance, disruption management, fuel optimization, crew planning and passenger rebooking. Hospitality companies could focus on demand forecasting, labor scheduling, energy management, preventive maintenance, food and water waste reduction and revenue optimization. These applications address systemic industry challenges more directly than simply adding to a growing list of AI pilots.

A hotel that knows a guest’s preferred room temperature may improve personalization, but an AI system that reduces energy consumption across thousands of rooms could produce a more meaningful financial and environmental benefit. Similarly, an airline virtual assistant may lower contact center costs, but an AI system that helps prevent cascading delays or rebooks disrupted passengers efficiently can protect both revenue and customer trust.

From AI product to operating capability

The tension emerging is between AI as a product and AI as an operating capability. When AI is treated as a product, companies may measure success by the number of pilots launched, interactions automated or employees trained. These metrics can create the appearance of progress without demonstrating business value.

When AI is treated as an operating capability, the starting point is a clearly defined problem: reducing delays, improving recovery times, increasing room profitability, lowering emissions or raising employee productivity. The measure of success then becomes the outcome rather than the technology itself.

Measure value, not AI activity

Companies should evaluate AI initiatives against their full cost of ownership and compare them with simpler alternatives. Sometimes better data, improved processes, conventional analytics or traditional automation may deliver much of the desired benefit at lower cost. AI should be selected when its ability to interpret complexity, predict outcomes or adapt decisions to changing circumstances creates value that conventional tools cannot provide as effectively.

This does not mean airlines and hospitality companies should slow down innovation. It means being more deliberate about where AI can have the greatest impact. The organizations that create the most value will not necessarily be those with the most AI features, but those that concentrate investment on high-impact problems with clearly established return on investment (ROI) measures. They will also need to retain human judgment where service, safety, accountability and empathy matter most.

AI adoption is becoming increasingly important to competitiveness, but it should remain a means rather than the strategy itself. In a sector facing tight margins and operational volatility, selective AI deployment is more likely to create durable value than broad, unfocused experimentation. The goal is to identify where AI can materially improve business performance and invest accordingly.