Google won an auction last Friday to acquire nearly all of Spirit Airlines’ employment and workplace records, a dataset that includes employee emails, chat logs, and performance files but excludes customer data. The tech giant agreed to use a court-appointed ombudsman to strip personally identifying information before the transfer, and promised never to intentionally re-identify the data. But the Association of Flight Attendants, which represents Spirit workers, isn’t buying it.
The union filed a court objection Tuesday arguing that Google’s privacy protections rely on consumer protection laws — and those laws don’t cover worker confidentiality. “The privacy architecture of this transaction is consumer-facing; its payload is disproportionately employee-facing,” the AFA wrote. That’s a sharp framing: Google built safeguards for a category of data it isn’t actually buying, while the data it is buying gets far weaker protection.
Former flight attendants are now rushing to object, worried that confidential information in their workplace records could be linked back to them by Google or a third party that later licenses the dataset. The concern isn’t hypothetical. Even de-identified workplace data — emails discussing medical leave, performance disputes, or union activity — can be re-identified through patterns that consumer protection laws were never designed to catch.
This is the same tension that has dogged Google’s healthcare AI deals with Ascension and HCA, where de-identification promises collided with worker and patient skepticism. The difference here is that the data isn’t medical records or purchase history — it’s the internal communications of people who had no meaningful choice in whether their employer handed it over. Whether Google can convince a bankruptcy judge that its consumer-grade privacy framework is enough for employee data may determine whether the deal survives at all.
