Google selected and financed the company that is anonymizing the data for Spirit Airlines.
Google has consented to pay $10 million for the knowledge held by a defunct airline. A judge is set to consider the sale on Wednesday morning, with Reuters reporting the price on Monday.
Every account of the agreement hinges on a single term: the data is deidentified, so there is no cause for concern. The data sale agreement involving Spirit Airlines, filed with the bankruptcy court on August 14, outlines three aspects of the process that have not been previously reported. When viewed together, these details alter the perceived value of the term.
Google selects the firm responsible for data scrubbing
Media reports have described an independent third party tasked with cleaning the data before it is transferred to Google. However, the contract is more precise. Spirit must provide the materials to one or more third parties that are “acceptable to or designated by Buyer.” Essentially, Google selects the agent.
Moreover, Google covers the expenses. The agreement designates the buyer as solely responsible for all costs associated with deidentification, specifying that these expenses do not impact the purchase price. Thus, the $10 million is simply a headline figure, not reflective of the actual bill.
Following this, Google has the right to inspect the work. Spirit is obliged to provide the buyer a reasonable chance to review and comment on the process and must give good faith attention to those comments. The standard requires that the scrub meets reasonable satisfaction from Google.
None of this is deemed improper; it is conventional commercial practice. However, it does not align with the typical impression of an independent audit.
The scrub must maintain record links
This is the critical clause, and it is one that has not been widely referenced. The agent is required to certify the work in accordance with the California Consumer Privacy Act standard. Additionally, health-related material must comply with federal health privacy regulations. Both regulations apply regardless of whether they would typically extend to this data, providing real protection.
The statement concludes with a stipulation: the certification must remain valid “while preserving referential integrity across the data set.”
Referential integrity implies that the links between data remain intact. This connectivity is intentionally preserved. A pseudonymous individual continues to be traceable from an email to a support ticket, a code commit, and a payroll record.
This is precisely what renders the archive valuable for training agents as it provides context, yet it also makes any anonymization potentially delicate, and the contract mandates this continuity.
What is contained in the data package
The detailed inventory of Spirit Airlines’ data is more extensive than previous summaries indicated. It outlines 100 million emails spanning 80,000 accounts, along with 500 million Teams messages. In addition, there are 17,082,644 OneDrive files, 20,577,677 SharePoint files, and 667,563 IT tickets. James Nani first reported these high figures for Bloomberg Law.
The engineering aspect includes 516 repositories and approximately 30 million lines of code, also documenting 372,585 commits and 43,170 pull requests, along with related pipeline logs.
The operational data is substantial—covering 763,391 flights and 5,014,676 crew pairings, it includes 190,312,864 booking records and 7,510,221,520 transactions dating back to May 2008. Entries related to disruption and reaccommodation add another 3,000,347,472 rows.
Furthermore, the corporate documentation encompasses board presentations, budget analyses, deal pipelines, due diligence reports, investment committee documents, lender materials, and merger fairness opinions.
Regarding the staff
The inventory details 175,658 employee records, with the record-keeping system dating back to August 1986. It also includes 3,426,618 payroll records and 148,018 employee tax forms, in addition to 1,092,000 time cards, training records, recruiting files, and travel requests.
Approximately 17,000 individuals were laid off when Spirit ceased operations on May 2. Their correspondence, pay histories, and tax documentation are now categorized as line items on a schedule.
These personnel signed employment agreements, not data licenses. In a Chapter 11 estate situation, this distinction becomes irrelevant.
What Spirit retained, and what it may still sell
The inventory also notes customer-related data that was not previously included. The quantities are substantial: 97.5 million customer profiles, 50.2 million Free Spirit members, 740,000 cardholders, 30,865,471 recorded calls, and 15,784,473 chat sessions.
Regulatory records are similarly categorized. This includes information on 2,491,715 disability service requests, data on denied boarding, and complaints to the Department of Transportation.
One clause merits close attention: Spirit is prohibited from selling the assets to anyone other than Google, with one exception: it may sell its customer data list, which contains individual traveler spending aggregated annually, to buyers in the hospitality or travel sectors.
Thus
Other articles
Google selected and financed the company that is anonymizing the data for Spirit Airlines.
The agreement for the sale of data by Spirit Airlines states that Google assigns the deidentification agent, covers the costs, and that the data scrubbing must maintain referential integrity.
