Expedia has articulated a set of core principles for building artificial intelligence at scale, drawing on years of deploying predictive systems across travel booking, fraud prevention, and customer support. The company warns that velocity without discipline is a liability.
Drawing from experience with billions of predictions, Expedia's framework addresses the shift from simple prediction to autonomous, agentic systems that can converse, reason, and take action on a traveler's behalf. The principles focus on reliability, governance, and accountability.
The travel giant has applied AI and machine learning across personalization, ranking, recommendations, fraud prevention, customer support, and generative AI experiences. Its insight is that building systems that continue to work, scale beyond individual teams, and improve consistently over time is harder than getting a model to work once.
As AI agents begin making decisions autonomously — such as booking travel on a user's behalf — the expectations around system reliability and accountability become fundamentally different. Expedia argues that the principles behind how these systems operate matter more than ever in this new paradigm.
The distinction between AI that works today and AI that lasts at scale is critical. Many companies optimize for short-term performance without considering long-term sustainability, a gap Expedia's framework aims to address.