Southwest Airlines uses generative AI in marketing and to modernise old crew software
Southwest Airlines' 2025 annual report says it introduced generative AI copy testing to improve marketing conversion and is expanding its use of AI and machine learning under an internal AI governance team. Separately, PwC used generative AI to turn the code of Southwest's old crew leave system into requirements, halving planning time by its account.
The operating picture
AI disclosed to investors
Southwest Airlines 737-800 jet. Photograph: Iflysouthwest / CC BY-SA 4.0 (cropped to fit). Illustrates the airline, not an AI deployment.
Southwest's annual report for 2025, filed on 5 February 2026, describes its AI use in some detail; Skift noted that Delta, United and Southwest are all disclosing more about AI in these filings. Besides marketing, it lists AI support for customer service replies among the year's technology work.
The filing also says the company has set up an internal AI governance team, with the board overseeing AI risks through its audit committee.
Selected implementations
The work behind the evidence
01 / Deployed
Generative AI in marketing
The operational problem
An airline that sells largely through its own website and app depends on being found in search, and on marketing copy that turns visits into bookings.
What the evidence reports
Southwest said in its 2025 annual report that it continues to invest in technology for marketing based on AI. In 2025 it optimised its app, landing pages and destination and market pages for search engines and generative AI engines, and "Generative AI copy testing has been introduced to refine marketing messages and optimize conversion rates, contributing to increased sales per marketing effort." The report says the company is expanding its use of AI and machine learning to carry out elements of its business strategy.
Operational significance
Optimising for generative AI engines reflects a shift in how travellers search for flights, and the annual report ties the copy testing to sales.
Evidence still needed
The report gives no figures for the increase in sales per marketing effort and does not name the tools used.
A source can verify an announcement without verifying its projected outcome.
02 / Reported · unverified
Reading old code to plan its replacement
The operational problem
Replacing an old crew leave and attendance system starts with working out what it actually does, which can take months when the documentation is thin.
What the evidence reports
PwC said in a March 2025 case study that it used generative AI on Microsoft Azure to analyse the source code of Southwest's legacy crew leave and attendance system and produce functional requirements, turned into a prioritised backlog. Backlog creation fell from 10 weeks to five, and 90% of more than 600 user stories written by the AI were accepted. Marty Garza, vice president of air operations technology, said: "With PwC, we condensed what could have been a months-long effort into a five-week milestone."
Operational significance
Many airlines run critical crew and operations systems on old software, and AI that documents them could shorten the path to replacing them.
Evidence still needed
The figures are PwC's. The case covers planning for the new system, not the system in operation.
Sources and verification (2)
PwC case study - Southwest Airlines (dated copy) · Evidence grade B
A source can verify an announcement without verifying its projected outcome.
The wider register
Other tracked activity
Other records are shown separately, including non-AI technology and programmes at different stages of deployment. A register entry does not establish a measured AI benefit.
AI-based support for service replies at Southwest Airlines · Deployed
A source can verify an announcement without verifying its projected outcome.
Editorial assessment
Southwest's documented AI is in marketing and in modernising its own technology, disclosed with unusual formality in its annual report.
The filing says AI use is expanding; evidence of where it reaches the operation would be the next step.
Live data from the Aerostack register. Status reflects published evidence only. See Our approach for how records are graded.
Aerostack register · 3 tracked deployments
Aerostack tracks AI deployments across airports and airlines, with the published evidence behind each record. Photography is real aviation imagery. Record thumbnails may show the named airport or airline; other photographs do not.
