Google Buys Spirit Airlines Data for $10 Million to Help Train AI Models
Google has acquired a large collection of internal Spirit Airlines data for approximately $10 million, giving the technology company access to millions of business records that could potentially help improve its artificial intelligence products.
The deal comes after Spirit Airlines shut down operations in May following 34 years in business. The airline had approximately 17,000 employees when it ceased operations and entered bankruptcy proceedings.
According to court documents and reporting surrounding the sale, the dataset contains an enormous amount of internal company information. The records reportedly include roughly 100 million emails, 500 million Microsoft Teams items, 17 million OneDrive files and more than 20 million SharePoint items. Other materials include documents, spreadsheets, calendars, payroll records, employee information and tax forms.
Google says the information it receives will not include personal customer data. Nearly 100 million passenger profiles and approximately 50 million records associated with Spirit's Free Spirit loyalty program are excluded from the transaction.
The company also says the data will be de-identified before being transferred. Spirit has indicated that a third party will handle the process and certify that the information being sold complies with applicable privacy laws.
However, the transaction has raised concerns from the Association of Flight Attendants-CWA. The union has filed an objection with the bankruptcy court, arguing that removing names and other identifying information may not completely eliminate privacy concerns.
The union specifically warned that confidential information could potentially be reconstructed, particularly when dealing with records involving large groups of employees or smaller identifiable groups.
Google's purchase highlights the growing value of real-world business data in the artificial intelligence industry. While synthetic data can be generated to help train AI systems, companies are increasingly interested in operational information that reflects how businesses actually function.
A competing bidder, Mercor, reportedly also pursued the Spirit dataset, demonstrating the demand for proprietary corporate information among companies developing AI and other technologies.
For Google, the purchase offers access to a massive collection of workplace data without relying on information obtained without permission. The company has said the dataset could be useful for improving its products and AI models.
The bankruptcy court is expected to consider the concerns surrounding the transaction at a hearing scheduled for Sept. 9.