Spirit Airlines Data Sale to Google: Unions Secure Privacy Protections

What happened

Spirit Airlines' planned sale of its data to Google has cleared a hurdle, and labour unions have secured privacy protections that will apply before any transfer. If the sale completes for $10 million, Google would receive about 1.9 billion Spirit emails and 30 million lines of code, which it could use to train its AI models.

According to Skift's reporting dated 10 November 2026, the unions' win means employee data will be removed first. The same reporting frames the agreement as conditional: the sale has not yet completed. The material available does not specify which body cleared the hurdle or when the transaction is expected to close.

Why it matters

This story shows how quickly accumulated communications can be recast as AI training material. An archive of billions of emails almost certainly contains personal information about passengers and staff, so the transaction raises questions about what people were told when they wrote those messages, and whether they would expect them to be used this way. For experience leaders, the lesson is that data governance is now part of the customer relationship, not just a legal back-office function.

The union agreement also matters as a precedent for how employee interests can be built into data decisions. Requiring that staff records be scrubbed before transfer shows that consent and privacy conditions can be negotiated at the point of sale, rather than discovered afterwards. Organisations that hold large volumes of historic correspondence should expect similar scrutiny from employees, customers and regulators.

By the numbers

• $10 million is the reported price for the proposed data sale to Google.
• About 1.9 billion Spirit emails would be included in the transfer if the sale completes.
• 30 million lines of code would also pass to Google under the same proposed deal.
• 10 November 2026 is the date of the Skift report covering the development.

The Renascence take

Most readers will focus on the $10 million figure, but the price is the least important detail. The real signal is how readily a company's historic communications can be valued as raw fuel for machine learning, with the people who created that data often left out of the decision. Protections negotiated after the fact show that privacy should be designed in from the start, not added as a final concession.

The headline price is a distraction; the operating question is what your data would look like to a buyer, and whether your customers would recognise that use as fair. Leaders should map where their emails, transcripts and support records sit, who can access them, and what they were told at collection. Ask whether a sale or AI training run would still feel acceptable if a customer or employee read about it in the press. If the answer is no, the problem is already in the experience, not in the contract.