Sabre's Agentic Blueprint: A practical guide to building AI in travel
At a glance
- 52% of executives report their organizations have deployed AI agents in production, per September 2025 Google Cloud study; agentic AI tools grow 35–46% annually.
- Sabre migrated petabytes of data to Google Cloud, enabling agentic-ready APIs and Model Context Protocol server for autonomous travel solutions.
- Trust emerges as primary adoption barrier; crawl–walk–run framework progresses from human augmentation through supervised automation to managed autonomy.
- Simple, granular APIs outperform complex monolithic structures; AI agents need domain-specific context and error handling to operate reliably in travel workflows.
What the report covers
Sabre's white paper outlines a framework for implementing agentic AI in travel, drawing on R&D, partner feedback, and industry research. It addresses technical architecture—agentic-ready APIs and Model Context Protocol servers—alongside organizational change management. The report targets travel technology leaders, DMOs, hoteliers, and travel agents navigating AI adoption, published December 2025. It is grounded in experiments with partners including Flight Centre Travel Group, Internova, and EF Education First.
Key findings
Sabre's research with internal teams and industry partners identified three critical barriers to agentic AI adoption in travel. Complex, deeply nested JSON payloads confuse language models and reduce reliability; simpler, granular APIs perform significantly better. One capability-rich API failed in testing due to overly complicated response structures, demonstrating that AI agents function more effectively with small, confident steps rather than monolithic endpoints.
Large language models lack built-in understanding of travel industry conventions—ticket types, PNR structures, interline agreements, fare rule nuances. Without rich API descriptions, examples, and defaults, agents make unreliable decisions. Error messages designed for human developers provide no path for AI self-reflection or retry logic. Domain-specific context embedded in API schemas is essential for agents to produce trustworthy outcomes.
Trust is the single largest barrier to autonomous agent adoption. Organizations fear uncontrolled autonomy, especially in customer-facing or write-access scenarios. The non-deterministic nature of language models makes debugging complex; identical queries can produce different outputs. Sabre partners—Internova, EF Education First, Flight Centre Travel Group—prioritize human-in-the-loop validation for complex workflows, viewing AI as co-pilot to empower advisors rather than replace them.
Sabre proposes a crawl–walk–run framework: Crawl stage uses AI for augmentation (parsing emails, summarizing policies, read-only queries); Walk stage introduces supervised automation via Model Context Protocol, where agents plan but humans approve; Run stage enables managed autonomy within pre-approved workflows with robust logging and guardrails. This phased approach transforms skepticism into adoption by building confidence incrementally.
Industry momentum supports agentic adoption: Forbes reports 35–46% annual growth in agentic AI tools; Google Cloud's September 2025 study finds 52% of executives have deployed AI agents in production, with early adopters seeing measurable ROI and revenue growth. Sabre's migration to Google Cloud infrastructure and deployment of agentic-ready APIs with MCP server positions the company to accelerate scalable adoption across travel sector.
Key numbers
| Metric | Value |
|---|---|
| Executives reporting AI agent deployment in production | 52% |
| Annual growth rate of agentic AI tools | 35–46% |
| Data migrated to cloud by Sabre | Petabytes |
Figures as published in the source; forecasts and survey results are labelled as such in the note.
Why it matters
DMOs & destinations
DMOs can leverage agentic AI to deliver personalized, real-time travel experiences—autonomous assistants that recommend nearby attractions, rebook travelers when plans change, and handle multi-step workflows without manual intervention. This builds visitor satisfaction and reduces operational friction. Adopting Sabre's agentic-ready framework positions destinations to compete on service innovation and customer experience quality.
Hotels & hospitality
Hotels benefit from AI agents that automate ancillary sales (checked bags, upgrades), handle booking changes, and provide policy-compliant rebooking options. This reduces customer service workload, accelerates revenue capture, and improves guest satisfaction. The crawl–walk–run adoption model allows properties to build trust gradually, starting with read-only queries before granting write-access autonomy, lowering implementation risk.
Travel tech & distribution
Travel-tech platforms must invest in modular, semantically rich APIs to enable AI agents to discover and use tools reliably. Complex monolithic endpoints obstruct agent reasoning; granular, domain-specific designs accelerate adoption. Platforms adopting Model Context Protocol and agentic-ready standards align with emerging AI tech stacks, reduce debugging costs, and unlock new revenue streams via autonomous booking and upselling workflows.
Methodology and limits
Sabre conducted internal R&D sprints testing prototype agents on existing APIs, revealing technical bottlenecks. Qualitative feedback was gathered from industry partners: Flight Centre Travel Group, Almatar, ACI Blueteam, EF Education First, Internova, and Direct Travel. The report synthesizes these experiments, partner roadmaps, and published industry research (Forbes, Google Cloud study from September 2025). This brief is based on publicly available summary text; the full white paper may contain additional detail. No quantitative survey sample size or statistical margins of error are disclosed in the available text.
Official source
The report is © Sabre. This brief is an original editorial summary by TourismIntel — it quotes only figures published in the source and never reproduces the document.
Check the official statistics on Pulse
Related reports
Sabre's white paper examines how autonomous AI systems transform security and governance requirements in travel technology.
Sabre's October 2025 whitepaper examines how agentic AI and conversational commerce will reshape travel retail globally.
Sabre's 'The Direct Connect Disconnect' report reveals a significant gap between industry ambitions for direct travel distribution and operational reality.
Sabre's white paper examines how proliferation of booking systems, data sources and API integrations has created operational friction in travel distribution.