Skyscanner's CEO on Why Agentic Booking Won't Kill Metasearch

The prevailing fear in flight metasearch is that AI agents will cut search platforms out of the booking flow entirely. Skyscanner’s CEO Bryan Batista argues the opposite—that agentic AI makes Skyscanner’s data more valuable, not less.

What Happened

Batista articulated Skyscanner’s agentic strategy in a recent interview: the company is systematically embedding its flight search logic across AI ecosystems. Skyscanner was a launch partner for OpenAI’s Operator agent in early 2025, integrated with Microsoft Copilot Actions, and launched a dedicated ChatGPT app. The pattern is deliberate multi-platform presence—no reliance on any single AI distribution channel.

On the booking question specifically: Batista said Skyscanner could build an agentic experience that completes bookings end-to-end, but that’s not the company’s strategy. “We are a search player, we don’t want to become an OTA.”

Why It Matters

Batista’s vision centers on agent-to-agent interaction as the paradigm worth designing for. In his framing, a traveler’s personal AI agent pre-loads preferences—fare types, cabin classes, loyalty memberships, preferred stopovers—and passes that context to Skyscanner’s agent, which returns hyper-precise results rather than a grid of options. ChatGPT’s “Atlas” browser is described as the first glimpse of this model in practice.

The strategic insight is that large language models are not trying to replicate Skyscanner’s aggregation work. “They’re just looking to provide the best answer possible,” Batista said. “We think we can play a big role in that ecosystem because these LLMs are not looking to contract with airlines.” Skyscanner’s moat is not the interface—it’s the pricing data, inventory coverage, and search logic that LLMs need to give correct flight answers.

This is a meaningful reframe of the metasearch threat model. The traditional fear was that natural language search would disintermediate the comparison grid. Batista’s argument is that comparison data becomes the foundation that makes natural language travel planning reliable. An AI agent that books without Skyscanner’s underlying data gives confident but wrong answers—a worse outcome than no agent at all.

What This Means for Travelers

For travelers who use AI assistants to plan trips, the agentic travel stack is taking shape through integration rather than disruption. The practical implication: flight search embedded in AI assistants will increasingly draw on Skyscanner’s pricing and availability data, even if the user never sees the Skyscanner interface. The quality of AI travel recommendations will correlate with how well platforms like Skyscanner have exposed their data to LLM-accessible APIs.

Source: Skift

Content created with AI assistance and reviewed for accuracy.