Lawmakers press Google to strip employee records from its $10 million Spirit Airlines data purchase

Spread $10 million across the inventory PYMNTS listed and Google paid about 1.6 cents an item ^[11]. The count is 100 million employee emails, 500 million Microsoft Teams messages, 17 million OneDrive files and 20 million SharePoint items, or 637 million in all ^[2][10]. On top of that come 516 code repositories holding 30 million lines of custom software ^[2]. Put the whole price on the code instead and it comes to about 33 cents a line ^[12]. In August PYMNTS reported that AI companies value this kind of proprietary, "real-world" data for training large language models on complex business tasks ^[5].

Emails and Teams messages make up about 94% of those items ^[13]. Horsford's office put the number of signers at 119, PYMNTS reported ^[14]. The CNA report says more than 120 ^[15]. The letter asks the companies to "exclude employee information from the transaction to the greatest extent possible" ^[4]. "Innovation cannot come at the expense of workers' privacy, and no employee data should move until real, enforceable safeguards are in place," Horsford said ^[9].

The privacy ombudsman's condition was narrower. On Oct. 5 the ombudsman recommended that the bankruptcy judge approve the sale at an Oct. 14 hearing, provided personal information from Spirit customers is excluded, Reuters reported ^[8]. That condition covers customers, while the letter is about former employees ^[1].

A Google spokesperson responded to both. "As we've said from the beginning of the process, we are not looking to buy any personal information from Spirit. The information will either be completely excluded or will be deidentified by an independent third party before Google receives any data. We're already working constructively with the appointed privacy ombudsman," the spokesperson told PYMNTS ^[7]. Deidentified records still reach Google. Excluded ones do not. Neither report says who pays the third party or what the work costs.

On Oct. 14 the judge could approve on the ombudsman's customer-only terms and leave employee records to Google's deidentification promise. A tougher order would exclude employee information outright. Then the cuts would fall on the 600 million emails and Teams messages that make up most of the purchase ^[13]. A negotiated version could sit between the two: deidentify the routine messages and pull the medical and disciplinary files the letter lists ^[6].

I'd expect the first outcome. The appointed ombudsman has already recommended approval ^[8], and Google says it has offered deidentification from the start ^[7]. If that holds, the letter costs Google a compliance pass of undisclosed size on a $10 million purchase. The Oct. 8 report on CNA still gives the price as $10 million ^[15]. So far, then, the risk of buying data out of a bankruptcy shows up as a compliance step, and the price has not moved. The counter-case is the letter's own list. A judge could rule that medical records and accommodation requests must be removed entirely, with deidentification allowed only for the rest ^[6]. An order on Oct. 14 that strips out employee records wholesale, or a reported cut to the price, would prove this view wrong.