Google Wins $10M Bid for Spirit Airlines' Data to Train AI

Alphabet's Google agreed to pay $10 million for a scrubbed corporate dataset from bankrupt Spirit Airlines, adding roughly 100 million emails and 500 million Microsoft Teams chats to its AI training pipeline.

Google Wins $10M Bid for Spirit Airlines' Data to Train AI

Alphabet's Google has agreed to pay $10 million for a large trove of internal operational data from bankrupt Spirit Airlines, adding roughly 100 million employee emails and 500 million Microsoft Teams messages to its AI training pipeline.

What Google is buying

The package, disclosed in a US bankruptcy filing, spans emails, Teams chats, spreadsheets, calendars, HR and financial records, marketing materials, project management documents, audits, pricing data on more than seven billion competitor flights and around 7.5 billion anonymized passenger transaction records collected over nearly two decades of operations. Google's winning offer beat a $7.5 million bid from AI data broker Mercor.

Personal data excluded

Google said the material will be "rigorously scrubbed of any personally identifiable information by a third party before receipt." Excluded from the sale are 97.5 million passenger profiles and 50.2 million Free Spirit loyalty records. "We acquired part of an enterprise dataset from Spirit Airlines, which can be helpful in improving our products and AI models," Google said in a brief statement.

Spirit Airlines logo

Bankruptcy auctions become an AI data channel

Spirit ceased operations in May 2026 after a bankruptcy reorganisation collapsed. A US bankruptcy judge for the Southern District of New York is scheduled to rule on the sale this week. The deal underscores how AI developers are turning to bankruptcy estates for proprietary corporate data that captures real-world decision-making, coordination and problem-solving that public web crawls cannot supply.

The purchase adds to Alphabet's expanding AI infrastructure push, alongside multi-gigawatt data-center partnerships and its own model roadmap. It also sharpens a broader debate over corporate data provenance, as regulators and workers weigh what happens to the digital exhaust of firms that no longer exist.

Reporting based on coverage from Bloomberg Law, 9to5Google, Business Insider and Axios.

Category: Business & Deals

Tags: AI AI Training AI Agents AI Infrastructure

Related Articles