Google buys data from bankrupt Spirit Airlines for AI training

What happened

Google has paid $10 million in a bankruptcy auction to acquire a large trove of operational data from defunct carrier Spirit Airlines, for use in training its AI models, according to Bloomberg Law reporting cited by Computerworld. The deal covers roughly 100 million emails, 500 million messages, and Microsoft Teams data, alongside records on revenue, flight operations, marketing, personnel and project management.

The package also includes some 30 million lines of program code, development data, and a range of software models and algorithms built by the airline over its years of operation. It further contains pricing data drawn from 7.2 billion competing flights and approximately 7.5 billion passenger transactions — a granular window into airline pricing and booking behaviour.

Google has said no personal data will be transferred as part of the sale. A third party will review the material first and strip out anything that could identify individuals before Google uses the dataset to improve its products and AI models.

Why it matters

The transaction is a reminder that a company's data estate can retain significant commercial value even after the business itself has failed. Corporate bankruptcies are increasingly generating datasets — internal communications, operational logs, proprietary code, pricing histories — that are attractive raw material for training large language models and other AI systems, precisely because they capture real-world, industry-specific behaviour at scale.

For technology and transformation leaders, this signals a widening market for "distressed data" as an asset class, with implications for how organisations classify, govern and eventually dispose of information assets, especially communications and pricing data that could later carry value — or risk — far beyond their original business context.

By the numbers

• $10 million paid by Google in the bankruptcy auction for Spirit Airlines' data
• 100 million emails included in the dataset
• 500 million messages, plus Microsoft Teams data, included
• 30 million lines of program code transferred as part of the sale
• 7.2 billion competing flights and approximately 7.5 billion passenger transactions of pricing data included

The Renascence take

Most coverage of this deal will focus on the price tag or the novelty of an airline's back-office data becoming AI training material. The more interesting story is what this says about the afterlife of operational and communications data — and how unprepared most organisations are for that afterlife.

Every email thread, pricing model and internal Teams message a company generates is, in effect, a behavioural record of how it actually served customers and made decisions — and that record doesn't disappear when the business does. The real lesson for experience and data leaders isn't about Spirit or Google specifically; it's that data governance can't stop at "in use." Organisations need a deliberate view on what happens to communications, pricing logic and proprietary models if the company is ever sold, wound down or acquired, because someone, eventually, will find value in it — and today it's increasingly an AI company doing the asking.