Airlines are not linking artificial intelligence (AI) to revenue growth, according to recent research.
An Accenture study suggested only 12% of carriers link AI directly to revenue, compared with other consumer industries that have adopted it more fully across customer-facing functions such as marketing, e-commerce and customer service.
The research, shared exclusively with PhocusWire, also showed generative AI is being treated as no more than an add-on. It revealed 58% of airlines layer AI onto legacy systems, rather than fully embedding it, which means its full potential is not unlocked.
The study claimed other consumer and retail companies are “weaving AI into the fabric of their businesses,” and said carriers could lag other industries unless they take a more strategic view of the technology.
A separate report from IT services company NTT Data revealed that the top AI-driven organizations are 2.5 times more likely to exceed revenue growth of 10% because of the “effective implementation” of AI solutions.
Emily Weiss, travel industry lead for Accenture, acknowledged that airlines have made progress in applying AI to customer-facing areas but said the greater opportunity is in transforming the core business.
“At present, too many airlines are still experimenting with small pilots or layering AI on top of legacy systems, which constrains scale and impact. To capture the full value, airlines must re-architect their systems with AI at the core, scale initiatives across functions and tie them directly to revenue outcomes,” she said.
But airlines have always been cautious regarding newer technologies because they are dealing with complex operations in a highly regulated environment with safety the priority. For those reasons, alongside slim margins, carriers are risk averse and concerned about disrupting existing systems.
The rearchitecture of systems is a huge undertaking but there is willingness. SITA research published in 2025 revealed gen AI as a top spend priority for 56% of airline IT professionals.
The report also revealed airlines’ desire to establish a “viable foundation for major technological innovations like AI.”
Airline technology companies agree airlines should be thinking about AI in terms of revenue growth. Airline retailing technology specialist Accelya described it as a “transformational” driver of revenue growth.
“When AI can be executed in real time, it has a measurable impact on fare yield, ancillary conversion and speed to market, with airlines already achieving fare revenue increases of up to 6% through AI-driven dynamic pricing,” said CTO Gaurav Roy.
According to Roy, there is a disconnect between the huge improvement in available insights and related recommendations driven by AI and the lack of action to turn them into revenue.
AI readiness
Roy said airlines should prioritize AI that can “actively price, construct offers, manage orders, service and resolve disruption in real time, within clearly defined airline guardrails.”
Others think airlines should move beyond pilots and prioritize AI readiness. A further finding of the Accenture research revealed that carriers are carrying out a handful of small pilots—14% reported six to 10 pilots—while almost half of other consumer-facing companies reported running more than 10 pilots.
“AI can only be as effective as the platforms it can access—it needs APIs that can handle its scale and order systems that unify customer data,” said Alex Mans, founder and CEO of FLYR, which also specializes in airline retailing and revenue management.
“AI that recommends a trip and a restaurant, but cannot price, book or service the order is a brochure, not a channel.”
Overall, technology players believe AI should be embedded across operations and customer experience so that carriers get the full benefit.
“The airlines that succeed will treat AI as essential infrastructure that strengthens reliability, differentiation and cost control over time,” said Simon Lomas, CIO of technology and engineering for SITA.
“Those that see it as a side initiative or a headline trend risk falling behind airlines that integrate it into how the business truly runs,” he said.
Roy highlighted examples, such as pricing and merchandising, where if AI was more fully integrated—acting on recommendations and adjusting pricing based on factors including demand and customer behavior—the return on investment would follow.
“The same principle applies to merchandising. AI can identify which bundles or upgrades will convert, but value is only realized when offers can be created, updated and deployed quickly,” he said.
Pricing concerns
It is worth noting that AI’s impact on pricing can be a sensitive subject. Many carriers are not public about their strategies or they tend to talk about AI deployments in terms of efficiency and optimization.
It could be in part because the area became a hot topic last year when Delta shared details of its revenue management technology partnership with Fetcherr.
The carrier said it was aiming to deploy the technology across 20% of its domestic network by the end of 2025. It also said it was “encouraged” by the early results of the AI intelligent pricing technology.
Not long after, a group of U.S. senators asked Delta for transparency on the data used to train the AI technology amid concerns of targeting consumers based on personal data. An official department of transportation investigation ensued and is ongoing.
Agentic moment
But putting the implementation of AI for pricing and merchandising aside, having AI as part of the infrastructure would also help carriers prepare for agentic AI, which is a “real moment of change in the relationship between airlines and travelers,” according to Mans.
He cautioned that airlines could lose customers if they cannot serve AI agents that are already helping customers plan trips.
Accenture said that airlines have barely begun their agentic AI journey, with 61% having no live projects.
Technology experts believe a step-by-step approach is necessary to build trust and confidence and target specific use cases where the technology can significantly transform operations.
Lomas said it should be a “teammate with wide visibility.”
“Early uses of agentic AI should support people, operate within clear guardrails, remain fully auditable and keep humans in the loop for operational decisions,” he said. “At its best, agentic AI becomes a digital coordination layer across the travel ecosystem, helping teams respond faster and work smarter, rather than replacing the people who keep operations running.”
Accenture conducted two surveys to assess maturity of airlines with AI (370 executives) and how they compare with other customer-facing industries (200 executives).