Google's $10 Million Spirit Airlines Data Purchase Triggers Employee Privacy Battle Over AI Training

Spirit Airlines Wants to Sell Its Data to Google. Former Flight Attendants Are Freaked Out

"It never crossed my mind that they would be so bold as to sell our private data for AI," says one former Spirit Airlines flight attendant. Spirit Airlines may have declared bankruptcy this spring, but the budget airline is still worth something -- to AI giants, at least. In mid-August, Google won a $10 million bid to purchase some 34 years of the airline's data, from invoices and flight operations information to Wi-Fi sales, employee records, and crew pairings. In a statement, a Google spokesperson said the data "can be helpful in improving our products and AI models." The sale would not include customer data, and Google "will not receive any personal information from this dataset," the spokesperson said. The winning offer, chosen over a competing $7.5 million proposal from AI data and training company Mercor, has to be approved by a judge. If former Spirit Airlines attendants have their way, that won't happen easily. Just days after the court announced Google's winning bid, the labor union representing 5,500 former Spirit Airlines flight attendants filed an objection to the sale. Lawyers for the 55,000-member Association of Flight Attendants (AFA) argue the sale would include an enormous amount of sensitive employee information, and that even Google's promised safeguards wouldn't prevent privacy violations of decades of flight attendants who never could have guessed that their data would be sold to train AI systems. The employee data "has no business being sold," Sara Nelson, the president of the AFA, said in a written statement to WIRED. "This is outrageous!" The legal objection opens a potential new front in the AI data wars, as major frontier labs including Google, OpenAI, Anthropic, and Mercor scramble to find new sources of data to help train their products. US laws have contemplated how best to protect consumer data, even after companies go bankrupt and are sold off for parts -- laws that have taken on new salience and value in the age of AI. But the Spirit Airlines' flight attendants' legal objection highlights the gap between data protections for consumers and for workers. It also marks the first public tangle between labor unions and corporations over the sale and use of employee data for AI training purposes, legal experts say. "There is no boundary between the information and data that the employee is producing and their own personal information. The law has not caught up," says Seema Patel, a law professor at the University of California, College of the Law, San Francisco, who studies labor issues and technology. (California is one of the few states to have explicit worker data protection laws.) "Companies are having a field day with this." In the past few months, startups specializing in selling defunct companies' data -- including old Slack messages, GitHub content, and Google Drives -- to those training AI have reportedly made millions. Meanwhile, so-called egocentric data collection efforts are on the rise, as companies race to record humans performing work like cooking meals, cleaning kitchens, and operating factory lines, and use that data to train machines to aid and eventually replace them. The Spirit data sale would, according to a court filing, include more than 1 million time-card records, over 175,000 employee records, nearly 150,000 employee tax forms, employment contracts and litigation files, 80,000 email accounts, 17 million individually owned Microsoft OneDrive items, 20.6 million shared Microsoft SharePoint files, and 500 million Microsoft Teams records. The court filing describes a process through which the data buyer -- Google -- would select or OK a third party to strip the data of elements that could be used to link the information with a particular consumer. A hearing related to the data sale has been delayed to September 9. One former Spirit Airlines flight attendant called the potential data sale "troubling." "I knew they were going to sell every single part of consumer data. It never crossed my mind that they would be so bold as to sell our private data for AI," says the flight attendant, who spoke to WIRED under condition of anonymity because they're still looking for new employment. The flight attendant says his and other employees' email accounts and Microsoft files included deeply sensitive personal and medical information shared with the airline as part of their employment and insurance agreements, detailing everything from miscarriages to incidents of domestic violence to union contract negotiations. "There's this huge consensus of feeling extremely violated by this," he says. In its legal objection, AFA argues that "deidentifying" employee data isn't the same as providing confidentiality, because AI itself makes it much easier to relink private information across datasets, even without specific names. "The privacy architecture of this transaction is consumer-facing; its payload is disproportionately employee-facing," the union's lawyers wrote. The union argues the bankruptcy court should either exclude all flight-attendant-related data from the sale or guarantee that employee data receives the same protections as consumer information. Though the flight attendants' objections are just a small part of the larger Spirit Airlines bankruptcy proceeding, the outcome could have a major effect on how labor unions and companies think about employee data going forward, says Ari Ezra Waldman, a professor of law at the University of California, Irvine, who studies privacy and technology. "In a world where billions and trillions are being spent on [large language models], anything where words are used is subject to its own profit-making purpose," he says. "If this goes forward with no limitation, there is no protection for any employee." The fact that flight attendants are at the forefront of this labor question, rather than workers who sit at desks and type all day, is a reminder that all employees can create data with market value. "All workers bring an immense amount of knowledge," Patel says. "Now artificial intelligence has given us the ability to process all different kinds of data." More value, meanwhile, is one of the former Spirit Airlines' flight attendants' goals. Separately, the union has also intervened in the bankruptcy to demand roughly $68 million in unpaid vacation, health care, and back pay. "We feel it's another slap in the face that we're waiting for our missing wages while Spirit is going to profit $10 million off our private information," says the former flight attendant.

Why Google Bid $10 Million for a Failed Airline's Data

What would a company like Google want with data from a failed airline? Google won a bankruptcy auction this month with a bid to spend $10 million on the corporate data of Spirit Airlines, which stopped flying in May. Google's offer beat a $7.5 million bid for the data from an AI data company, Mercor. The cache includes 100 million emails and 500 million Microsoft Teams messages, and around 30 million lines of code. AI companies have been battling it out this year to make the best AI coding agent -- and they have benefited from an abundance of publicly available code to train these models. But teaching agents to do other kinds of white-collar work may require data that is largely private, buried in corporate emails and chats. If that effort succeeds, it could send ripples across the labor market. Recommended Stories "Buying Spirit Airlines -- that's not centrally a coding agent purchase," says Nick Heiner, the head of reinforcement learning environments at Surge, a data company that supplies the frontier AI labs. "That's a purchase if you believe that your agent is going to generalize into the rest of the economy." Reinforcement learning One of the biggest trends in AI over the last 18 months or so has been the rise of reinforcement learning from verifiable rewards -- a method of AI training that has led to rapid improvements in models that can work independently for long periods of time. Anthropic alone has discussed spending more than $1 billion in a year on so-called reinforcement learning (RL) environments, which are essentially simulations of commonly used software, where AI agents can take strings of actions independently. If these actions lead to a desired outcome, the AI is rewarded and will be more likely to take similar actions in the future. Until now, the biggest jumps from this type of training have come from coding models, mostly because code has a useful property: it either works or it doesn't, meaning that the reward signal is immediate, so improvement can happen in a fast loop. (It's also helpful that there was plenty of coding data already out there on the internet, meaning models were good coders to begin with.) But AI companies' long-term goal is to automate large swathes of the economy. That's where Spirit's data likely comes in. What makes the data useful RL environments are only as good as the data that populates them, says Heiner of Surge AI. Companies like Surge and Mercor often hire human workers who are tasked with populating these environments with realistic data, either from scratch or in partnership with AI tools. "But even that is a little bit removed from literally having actual data that was used in the real world," Heiner says. "That's where deals like Spirit come in." Even though Spirit was a failed company, its data probably still encodes many useful things about how humans carry out their jobs, like how workers collaborate toward shared goals, or the norms of how the aviation industry operates. Even data depicting bad decisions can be helpful, Heiner says. "Maybe you see something where they're actually doing something poorly, and you build a task out of that, and you change what they were doing to be something better." But unlike coding, which can be quickly and easily determined to either work or not, office tasks are fuzzier. Human experts can still provide that feedback -- a service that companies like Mercor and Surge offer to AI companies -- but that work is relatively expensive, time-consuming, and subjective. It's therefore unclear that even populating RL environments with real-world data like Spirit's will allow AI companies to move into white-collar fields as quickly as they have ripped through the software industry. Privacy drama A Google spokesperson declined to comment on the specifics of how Spirit's data would be used, but said the data "can be helpful in improving our products and AI models." The spokesperson added that personal information would be "rigorously scrubbed" by a third party before the dataset's delivery to Google, and would include no customer data. After Google submitted its bid, a union of flight attendants filed an objection, saying that although the deal's terms would result in personal data being removed, the dataset would maintain "referential integrity," meaning that links between one type of data and another would be kept. That integrity is what makes the data useful for training AI, but the union said it may also allow for even anonymized information about its members to be reconstructed. The group is seeking additional privacy protections; a judge is set to rule on the matter on Sept. 9. Another AI data company, Micro1, submitted a higher bid of $12.5 million after the auction closed. Whichever way the deal goes, it looks like the AI industry is setting its sights on white-collar work. "Google is willing to pay $10 million for a failed budget airline, because if you believe this is really generalizable, you see a path to all these different industries -- not just the ones that Spirit was involved in -- being something you can replace pretty quickly," Heiner says.

Google Bought Bankrupt Spirit Airlines' Internal Data for $20M -- Now Experts Fear AI Could Replace Entire Corporate Teams

Google may have won its $20 million bid to acquire vast reserves of bankrupt Spirit Airline's business and employee information for use training its Large Language Models. But that doesn't mean the company and other tech giants won't continue facing pushback from critics who object to the tech sector using -- yet again -- third party individuals' data to develop artificial intelligence tools that increasingly risk automating other people's jobs away. Meanwhile, some opponents argue the sale of Spirit's data not only risks compromising its own employees' privacy, but may also eventually inspire some solvent businesses to consider selling their non-sensitive data to AI firms as a new income stream. That potential is even more worrying as some companies already begin using staff data to make apps designed to copy the work and production of the people they're replacing. Spirit's employee union files suit against internal data sale It didn't take long for the first salvoes to be fired at the deal that Google made last week to buy long-troubled Spirit, which finally went bust in May with more than $8 billion in debt. The first shot came Tuesday with a legal suit filed by the Association of Flight Attendants (AFA), which represents all the airline's former workers. While that case isn't seeking to block the sale of the company's data to Google -- a transaction with an AI firm some observers consider inevitable -- it does question the Silicon Valley giant's commitment to masking the identity and privacy of Spirit's staff. Doubtless anticipating just that kind of opposition, the sale of Spirit's internal business records and communication logs excluded data directly from, or even about its former customers. But it does cover the company's financial, marketing, HR, coding, and other records, including exchanges between employees on messaging apps, in emails, and even productivity reports. That, the AFA claims, is exposing Spirit's ex-workers to identity risks that former passengers are being spared. "Nearly every category under the heading 'Team Member,' time card information, employee data and employee records, employee business travel records, corporate and crew training records, payroll records, employee tax forms, and employee documents, is designated 'included,'" the filing argues. "The privacy architecture of this transaction is consumer-facing; its payload is disproportionately employee-facing. Hence, the employee data is far more confidential than the customer data, yet receives far less protection than the customer data." Google has agreed to allow a court-appointed expert to oversee scrubbing the "personally identifying information" in those records. But the AFA is concerned the company could eventually use its own tech to de-identify Spirit workers. That risk will increase, the AFA says, as the data inevitably comes into possession of other companies developing AI, and which either buy it from Google or obtain it by some other means. AI increasingly trained to run business units, not just a job Meanwhile, myriad other companies facing financial difficulties -- or simply seeking new forms of revenue -- could eventually also decide to sell tech firms the internal and employee information they don't consider critical to their businesses. But the non-profit digital rights organization Electronic Frontier Foundation (EFF) argues neither solvent nor bankrupt firms should be allowed to market that data unless individual employees who generated it agree.

From Spirit Airlines' graveyard: planes, headquarters -- and now a data sale

Spirit's liquidation has proceeded as you might expect for a bankrupt airline: planes sold or returned, landing rights auctioned to other carriers, the headquarters bought by another firm. Now the process is colliding with the times in which we live -- the digital age -- with conflict involving Google and other Silicon Valley companies. Massive troves of data that the Broward-based airline maintained, including about employees, are being put up for sale. They include at least 100 million emails, tens of millions of files stored in the cloud and tons of chats and calendar items. At least three California-based tech companies are involved in the dispute, each with different roles. Two of them want the data to train their artificial intelligence technologies. As hot as AI is, real-world data needed to train the machines is even more lucrative, and tech companies are willing to pay millions of dollars. But the auction process has angered the union representing Spirit's former flight attendants. They worry sensitive information about former employees will be made public, and that enough protections aren't in place. In filings with the New York bankruptcy court overseeing the case, both Spirit's representatives and the interested bidders say privacy and personal identity will be protected. They also say what's being sold is internal company data, not information about consumers. But those promises haven't soothed critics. At stake is not just concerns over privacy but also ownership of intellectual property, or who has the right to sell what. "All the employees that trusted the company with their information are at great risk," said Adam Schwartz, privacy litigation director at Electronic Frontier Foundation, a San Francisco-based non-profit focuses on digital civil liberties. "We urge companies like Spirit not to do this." On Aug. 14, Spirit and its representatives held an auction for large amounts of data the airline maintained. Google won with a bid of $10 million. Google agreed that the data chest would first go to a third-party firm that acts as a "deidentification agent." That's a company that can scrape off personally identifying characteristics. And the data awarded didn't include the personal information of customers or consumers, both Google and Spirit said. The "buyer publicly commits to maintain and use the deidentified data in deidentified form, and that it will not intentionally associate the deidentified data with any person or household," said a court filing. The data also wouldn't include "privileged materials," according to the filing. "We're acquiring part of an enterprise dataset from Spirit Airlines, which can be helpful in improving our products and AI models," Google spokesperson Gareth Evans said in an email to the Miami Herald. "We will not receive any personal information from this dataset. Any data we receive will be rigorously scrubbed of any personally identifiable information by a third party before receipt." The amount of data awarded in the auction is massive. Google obtained access to productivity and collaboration data, "including, without limitation, email data, calendar data, chat, collaboration and messaging data, documents, presentations, spreadsheets, knowledge repositories and wikis," according to court documents. Former Spirit employees don't seem to be protected. Spirit's representatives also picked an alternate bid, from San Francisco-based AI company Mercor for $7.5 million. But the union representing former Spirit flight attendants has criticized the carrier for selling the data. The Association of Flight Attendants-CWA, AFL-CIO filed an objection with the New York bankruptcy court, saying that former employees would be put at risk if confidential work information were released. "The proposed sale would transfer to Google LLC (and, in the alternative, to Mercor.io Corporation) substantially the entire employment and workplace record of Spirit Airlines, including decades of payroll, timekeeping, training, travel, and recruiting files," the Aug. 18 objection said. That's "together with approximately 100 million emails, 80,000 email accounts, 17,082,644 OneDrive items, 20,577,677 SharePoint items, and 500,000,000 Teams items." The union gave examples of its concerns, saying a flight attendant's disciplinary correspondence, a request for leave or an internal Teams exchange about staffing remain sensitive even if employee names are stripped. It said there's no protection against confidential information from being divulged. While the union acknowledged that consumers do get some protections, it wrote: "The employee data is far more confidential than the customer data, yet receives far less protection than the customer data." Other information Google would get access to comes from over 1 million time cards, 148,000 employee tax forms and over 3 million payroll records, according to the union. Whether removing identifying information is sufficient to protect privacy, Electronic Frontier Foundation's Schwartz said one could delete the "To" and "From" on an email but "what's left will nonetheless contain information that can identify someone." He said rather than companies making decisions on removing identifying information and what's sufficient, the employees should get to make those decisions. "It's their information and it's their privacy risk." Following the objection from the Association of Flight Attendants-CWA, the court postponed the hearing to approve the sale of the data from Aug. 19 to Sept. 9. For another Silicon Valley company, that delay became an opportunity, Micro1 missed the original court-imposed deadline for bids but contacted Spirit's representative last week, expressing interest in the data, according to a person familiar with the tech company's thinking. It now expects to make a bid of $12.5 million that will enter the court docket. The company is taking more time to prepare its offer, hoping to address the flight union's concerns. Ali Ansari, founder and CEO of micro1, declined to discuss specifics of their interest but said in an email to the Miami Herald on Monday that "we are moving quickly to prepare a bid for filing on the docket." He noted that "as part of that process, we plan to engage with AFA [the flight attendants union] to understand its concerns regarding the current transaction and explore whether our proposal could address those concerns and earn the union's support." That may prove challenging. Association of Flight Attendants President Sara Nelson told the Herald on Aug. 22: "Our data simply is not for sale. We did not and do not consent." She noted that Spirit flight attendants still haven't been paid their accrued vacation, sick leave and other overdue compensation. "Attempting to now sell their data is adding insult to injury. It's outrageous." There is other pushback, too. Spirit, Micro1 and Google have another objection to consider. On Aug. 22, a third company raised an objection to the sale to Google in a court filing it made with the New York bankruptcy court overseeing the case. But it wasn't trying to outbid the others. Its argument was that it has always been the rightful owner of much of the data and software Spirit used to maintain it. San Francisco-based Springshot said Spirit was a customer dating back to 2022 and that a court filing provided terms of the agreement. The tech company developed a software platform that analyzes and synthesizes tons of sources of operational data. It organized that information in ways that allow employees to know what tasks to complete and when and make quicker real-time decisions. That ultimately helps reduce flight delays. At the conference Future Travel Experience Global 2022 in Las Vegas, Mike Byrom, Spirit's then vice president of airport services, discussed the partnership with Springshot and its software. "We are deploying at every airport, at every gate, on every departure, and our ramp also has full visibility of this as well," Byrom said in a presentation on Sept. 8, 2022. The Herald viewed a video of the talk. Spirit used Springshot through its final flights on the morning of May 2, 2026. In a filing, Springshot requested that the court consider "Springshot's ownership of substantial intellectual property utilized by Spirit Aviation for its operations and which appear to be among the categories of data Google is seeking to acquire" and block the sale. Springshot said that it didn't wish to stop the general liquidation process but was calling for "a transparent process." The company said in a court filing that its customer agreement with Spirit said that Springshot "owns all right, title and interest, including all intellectual property rights, in and to the software, service, any data or information generated by the service or software." Springshot CEO Doug Kreuzkamp was more direct. "Springshot's intellectual property is not for sale," he told the Miami Herald. "The definition of the data set that Spirit seeks to auction off includes Springshot's intellectual property," Kreuzkamp said. "Spirit is not authorized to transfer Springshot's intellectual property." The company's top executive, born in Fort Lauderdale, acknowledged they missed the initial deadline the court imposed to raise objections. He said "that's because we had no reason to believe that Spirit would sell property that wasn't theirs." But since realizing what transpired, they've moved as quickly as they could, he noted. In a court filing, "Springshot respectfully requests that the court decline to approve the contemplated transaction unless Springshot's property is expressly excluded from the sale."