Much of the recent debate around artificial intelligence (AI) in travel has been about how far online travel players might be disintermediated.
Consumer research, alongside an audit of the AI platforms and initiatives from the online travel agencies (OTAs), reveals the greater impact could be on their margins.
The study of leisure travelers in the U.S. from investment bank BTIG, conducted June 26, has found significant consumer willingness to use the available AI tools for inspiration, planning and even booking.
More than 60% of travelers have dabbled with AI, with the majority comfortable using it across inspiration (76% very or somewhat willing), research (80% very/somewhat) and itinerary building (65% very/somewhat).
These figures are consistent with a trend Phocuswright has also seen in its research. In its most recent survey of U.S. travelers in February, 56% had used AI for planning, booking or in-destination assistance for at least one trip in the past 12 months, up from 43% in the second half 2025 and 33% in the first half 2025.
BTIG's study of 250 U.S. travelers also found that 38% would use it to book a flight and 42% to reserve a room. However, the study concluded that the willingness to use AI tools is ahead of the current AI trip planning functionality.
BTIG, part of U.S. Bancorp, noted that "tools are rudimental" with little personalization, only an inkling of pricing and no booking functionality. Despite partnerships between OTAs and other travel players, the study said the landscape has not moved on that much from when it conducted a similar exercise in 2023.
"Our view continues to be that AI is most likely to evolve as a customer acquisition channel versus disintermediation threat with the OTAs positioned as partners of choice given decades of behavioral data, millions of wired properties and billions to spend on marketing," BTIG said in its note.
Sales and marketing spend in 2025 by the largest online travel players—Airbnb, Booking Holdings, Expedia Group and Trip.com Group—exceeded $20 billion. The figure was up from $17.8 billion in 2024.
Alongside the consumer data, BTIG looked at the current AI trip planning offerings of OTAs such as Booking.com and Trip.com and that of AI platforms including Google Gemini, OpenAI's ChatGPT and Anthropic's Claude.
It found OTA initiatives were either weak when it came to "inspiration-itinerary building or limited to AI-powered search." Travel startups in the space, meanwhile, leaned towards inspiration-itinerary building and were weak when it came to pricing, availability and booking.
The AI platforms are therefore probably "most relevant to the broadest cross-section of travelers," the report said.
"All were pretty good at surfacing destination information, things to do and itinerary creation, but we found no real personalization, underlying content skewed to OTAs (similar to traditional search), only indicative pricing and had no ability to complete a booking."
The topic and challenges of booking capabilities in AI platforms has been ongoing for some time. OpenAI deprioritized its Instant Checkout functionality in ChatGPT earlier this year, and Perplexity has sought partnerships with companies such as Selfbook. Travel executives also believe the future is likely some mixture of "interface-led, ecosystem-led and transaction-led."
The BTIG study positions OTAs as the way to bridge the gap between intention and functionality, with partnerships the way forward, and said the risk of disintermediation has been exaggerated.
"We're thinking the large platforms compete more against traditional search with things like price comparison sites more at risk than OTAs."
It added that AI platforms currently account for only about 0.4% of OTA traffic but questioned how far that traffic could eat into the roughly 40% of traffic OTAs receive from traditional paid search currently.
The debate will continue as executives from across the travel industry hold very different views on the winners and losers as AI bites.