Flight Attendants Challenge Google's $10M Spirit Data Deal Over Worker Privacy Gaps

Key points

• Google agreed to pay $10 million for Spirit's data, covering the cost of a third-party service to scrub personally identifiable information (PII) before transfer.
• The dataset includes employee emails, HR information, payroll data, productivity metrics, and internal communications, but excludes customer profiles and loyalty program data.
• The AFA argues that de-identification removes names but does not protect confidential employment details, such as disciplinary records or union grievances, which could still be inferred from the data.
• Google states it will not intentionally re-identify individuals and will use the data to improve its AI products, while the AFA seeks court approval to exclude flight attendant data entirely.
• Mercor Corporation was the next-highest bidder at $7.5 million but was rejected for refusing to use a third-party scrubber, instead offering to scrub the data itself.

Background

Spirit Airlines ceased operations on May 2, 2026, following its second Chapter 11 bankruptcy. Since then, its assets have been auctioned off, including a significant portion of its internal data. Previous coverage indicated that the sale was part of a broader trend of acquiring failed companies' internal data for AI training, with JetBlue separately acquiring Spirit's LaGuardia slots for $58.5 million.

How outlets are covering it

Ars Technica highlights the AFA's argument that the privacy architecture is 'consumer-facing' but the data payload is 'employee-facing,' creating a loophole where worker confidentiality is not protected by consumer privacy laws. Axios focuses on the commercial aspect, noting that Google will use the data for AI training and that the deal excludes passenger data. Both sources agree on the $10 million price and the exclusion of PII, but Ars Technica emphasizes the legal and ethical concerns raised by the AFA, while Axios notes the technical details of the de-identification process. The AFA's objection is described as 'limited' and does not seek to disrupt the sale, but rather to ensure worker protections are in place.

Why it matters

This case highlights the growing tension between AI training data needs and worker privacy in the context of corporate bankruptcies. It raises questions about the adequacy of de-identification in protecting confidential employment information and the role of courts in balancing commercial interests with privacy rights. The outcome could set a precedent for how employee data is handled in future bankruptcy sales and AI training initiatives.

What to watch

A federal judge is expected to rule on the sale in September, potentially after a hearing where the AFA's objection will be considered. If the sale is approved, Google will receive the de-identified data for AI training. If not, Mercor Corporation may step in as the backup bidder. The AFA has requested that the court deny approval until Google agrees to exclude flight attendant information and notify workers of third-party access.