United Airlines’ AI strategy: The airline that makes decisions fastest wins

At United Airlines, AI has been a long-term strategic investment, not a recent initiative. While many companies scrambled to adopt AI capabilities during the recent gen AI boom, United had already built the foundation necessary for effective implementation. This forward-thinking approach stems from a clear business philosophy that in the airline industry specifically, the carrier quickest to make complex decisions gains the competitive edge. United’s methodical building of data infrastructure, compliance frameworks, and specialized talent demonstrates how traditional companies can develop true AI readiness that delivers measurable results for both customers and employees.

How have you prepared United Airlines for the current state of AI innovation?

Prior to generative AI, we used AI for customer personalization and marketing campaigns, as well as in our contact centers to help agents deliver more personalized service. We’ve been building the gen AI foundation for years because we know the most decisive airline wins. The complexity could be customer distress, a storm, an airport slowdown, or any other situation with a lot of data and urgency to empower employees and customers with relevant, in-the-moment information. Much of this work has been in organizing our data and building a secure platform for machine learning and other AI modeling. We also built an organization skilled in the data engineering and data science required for AI.

What are the essential building blocks to get into a position of AI readiness?

The list is long, but I’ll focus on two key capabilities. The first, which is half the battle, is getting your arms around the data and making it available, which means having the engineering ability to abstract it for use in the models. The second is having connective tissue between the technology, operating, cyber, and legal teams to create a compliance structure required to deploy AI solutions with the proper safeguards. The cross-functional risk management team is also essential because you don’t want to jeopardize your entire business over an AI pilot.

Talk us through a gen AI use case.

The leadership team agrees that the ability to communicate to our customers in clear and transparent English is a top priority for AI investment. Years ago, before gen AI, we started a program called Every Flight Has a Story. If we had a tough day of flight cancellations, our team of storytellers would gather all the operational data, speak to the right people, read the notes, and then craft a message to our customers, in an agreed upon style, about the status of the flight.

This was amazingly effective, but the challenge was in scaling it. Considering the number of delays versus storytellers, we couldn’t have a person write a new message with every event. So we focused on prioritizing the most impactful situations. Our first real gen AI use case has been to scale our customer messaging.

So how did you manage all the data?

The data piece was simple: the basic facts of the flight and the running chat between the attendants, pilots, gate agents, and the operations people associated with the flight. We fed that information — with additional data on weather, for example — into the AI model, to generate a good draft customer message.

The trick then was to have it understand the nuances of United Airlines’ communications style and what we wanted to emphasize. That’s where prompt engineering came in, not to train the model to understand flight data, but to use the words United prefers. Let’s take safety, for instance. We can emphasize safety with without scaring people, and the AI tool is learning to make the right word choice.

We were excited to discover that in addition to learning how to craft the message, the AI model was very good at looking back in time to bring previous flight data into the current situation. Even our human storytellers didn’t include reasons for flight delays, and that kind of information can be very useful to a customer.

When we launched the model, we went from putting 5% of our messages through the full storyteller process to 35% today, a 700% increase in volume with the same number of storytellers.

What internal challenges has gen AI helped you to solve?

Managers are very good at knowing their team’s performance and development plans, but they often find evaluations to be time consuming and not always helpful to the employee. United GPT, our internal chat GPT, assists managers to write evaluations. The manager provides employee notes, the tool writes a draft, and the manager reviews and approves. The feedback has been overwhelmingly positive, and now more than 90% of our managers use it.

In terms of driving adoption, United GPT sold itself. We just gave people the link and the adoption rate spiked for a long time. With gen AI, we have a wow factor that makes driving adoption less challenging.

What skills are you developing to continue this gen AI momentum?

We’ll continue to need data engineering and analytics, data science, and prompt engineering. I won’t know how many people until I know which tools will have embedded AI and which capabilities we need to build ourselves. One question we’re wrestling with is I know we get a productivity boost when automation shifts us from writing code to accepting code, but I don’t know if that’ll reduce my need for developers. We’re in a world where customers demand more, so I might need even more developers than in the past. If everyone can go faster, they will.

Given the seemingly limitless possibilities of AI, the CIO role is changing. What’s your advice to those who might eventually have your job?

The critical factor for the CIO is the ability to articulate a future that’ll be very different from today, and that gen AI brings disruption, not a continuation. As CIO, accept the fact you can’t be exactly right about the future, but you should form and articulate opinions so you can move your organization in the right direction.

We have many examples of market leaders that didn’t see new competitors coming behind them. The CIO skill is the ability to help people anticipate the future so they can get the most out of it. By the time customer service changes forever, you’ve made sure your board, team, organization, and technology are ready.