In a Crisis, Travel Companies Count on Humans — Not AI

Why It Matters

The incident reveals that current AI solutions cannot reliably handle large‑scale disruptions, prompting a reassessment of technology reliance in travel customer service. It underscores the need for resilient, hybrid support models to protect brand reputation and passenger experience.

Key Takeaways

• AI promises fell short during Middle East crisis
• Human agents overwhelmed, causing long wait times
• Travelers formed informal networks for real‑time updates
• Only Air India Express, Expedia leveraged AI effectively
• Crisis reveals gap between tech marketing and operations

Pulse Analysis

The travel sector poured billions into AI‑driven contact centers after the pandemic, touting chatbots, predictive routing and cost reductions as the next competitive edge. Those promises gained traction as airlines and OTAs marketed instant, 24/7 assistance that could scale during peak disruptions. When the February 28 U.S.–Israeli strikes forced a blanket closure of Middle Eastern airspace, the true test of those systems arrived. Over 43,000 flights were cancelled, thrusting millions of passengers into a logistical nightmare that quickly exposed the disparity between AI hype and operational readiness.

During the shutdown, most carriers defaulted to traditional call centers, routing travelers to human agents whose queues swelled to hours. The few firms that kept AI in the loop—Air India Express and Expedia—found their bots unable to handle complex rebooking rules, visa concerns, and real‑time flight status changes, leading to dead‑end responses. As a result, passengers abandoned official channels, congregating in social‑media groups and messenger chats to share flight updates and coordinate evacuations. The episode highlighted that current AI models lack the contextual depth and decision‑making authority required in large‑scale crises.

The fallout forces travel executives to rethink crisis‑response architectures. A hybrid approach—combining AI for routine inquiries with empowered human agents for exception handling—can reduce wait times while preserving flexibility. Moreover, AI vendors must train models on rare, high‑impact scenarios and integrate real‑time data feeds from airlines, airports and regulatory bodies. Investors are now scrutinizing AI spend, demanding measurable resilience metrics rather than vanity KPIs. Ultimately, the Middle East airspace shutdown serves as a cautionary benchmark, reminding the industry that technology alone cannot replace the human judgment needed when flights ground en masse.