Airlines Turn to AI as Cold Snap Tests Operations

The severe weather currently affecting the US is placing considerable strain on the country’s airline industry, causing schedule and route adjustments that are impacting global travel.

During such events, airlines face a sharp rise in customer inquiries and must make rapid operational decisions—all while adhering to the strictest safety standards.

Several carriers are now adopting generative AI to manage these disruptions and improve overall efficiency and responsiveness.

Last year, Air France-KLM established a cloud-based generative AI 'factory' to streamline AI development across the organization, making it more consistent and reusable. In partnership with Accenture and Google Cloud, the airline tests and deploys generative AI models that deliver measurable results in ground operations, engineering, maintenance, and customer service. The collaboration has reportedly accelerated generative AI deployment by over 35%.

The AI factory builds on earlier work with Accenture to migrate core applications to the cloud. Since then, Air France-KLM has developed a private AI assistant and RAG tools that connect large language models with internal search, supporting tasks such as diagnosing and repairing aircraft damage.

Employees are also trained to use these AI tools, enabling them to leverage the power of LLMs to drive positive business outcomes.

How Airlines Are Using AI During Severe Weather

United Airlines is also integrating AI into its operations. In an interview with CIO.com, CIO Jason Birnbaum explained that AI helps "shorten decision cycles" during irregular operations, such as those caused by recent extreme cold weather. The airline first adopted AI to handle passenger inquiries.

When flights are delayed or canceled, customer service representatives must respond quickly and informatively while maintaining the company's approved communication style—refined through its 'Every Flight Has A Story' program. During prolonged disruptions, it becomes challenging for these "storytellers" to keep up.

Birnbaum noted, “Given the volume of delays compared to available storytellers, we couldn't have a person craft a new message for every incident. We focused on the most impactful situations. […] We fed the AI model with basic flight details, real-time communication among crew and ground staff, and supplementary data like weather conditions to generate a solid draft message for customers.”

“The challenge was ensuring the AI understood United’s communication style and key priorities. That’s where prompt engineering came in—not to train the model on flight data, but to adopt United’s preferred phrasing. For example, we emphasize safety without alarming passengers, and the AI is learning to choose words accordingly. […] The model excelled at incorporating historical flight data into current situations. Even our human storytellers rarely included reasons for delays—information that customers find very helpful.”

According to the Boston Consulting Group, the airline industry’s AI maturity is now 'average,' having improved from slightly below average over the past year. Of 36 airlines surveyed, only one met the highest readiness criteria for an AI-driven future. The analysis indicates that by 2030, carriers embedding AI at the core of their workflows could see operating margins 5–6 percentage points higher than their peers.

Generative AI is expected to become integral to airline and airport operations, supporting quick decisions on schedules, crew allocation, aircraft rotations, and passenger recovery. Microsoft states that data-driven AI systems can reduce the root causes of flight delays by up to 35% through better disruption forecasting, minimizing the ripple effects of travel interruptions.

Airlines using AI-driven personalization report revenue increases of about 10–15% per passenger, according to Microsoft. The company also highlights that AI-based self-service interfaces can cut costs by as much as 30%.

Interested in learning more about AI and big data from industry experts? Attend the AI & Big Data Expo in Amsterdam, California, or London. This comprehensive event is part of TechEx and co-located with other leading tech events. Click here for details.

AI News is powered by TechForge Media. Discover other upcoming enterprise technology events and webinars here.