Alphabet’s Google has agreed to purchase internal business data from the bankrupt Spirit Airlines for $10 million to train its AI models. The deal, which includes millions of corporate emails and internal messages, awaits a court hearing on August 19, 2026. This acquisition highlights the growing value of legacy corporate data in the artificial intelligence sector.
Alphabet Inc.’s Google has secured a deal to purchase a large trove of internal business data from Spirit Airlines, the carrier that ceased all operations in May 2026. The tech giant has agreed to pay $10 million for the archives, which it intends to use for product development and the training of its artificial intelligence models. The sale is currently pending final approval from a U.S. bankruptcy judge, with a court hearing scheduled for August 19, 2026.
The acquisition is notable for the scale and type of information being transferred. The dataset reportedly includes approximately 100 million emails, 500 million Microsoft Teams messages, software code, and various operational, marketing, and productivity records. For AI companies, this level of detailed corporate interaction is highly valuable, as it provides real-world examples of how large organizations communicate, manage logistics, and execute daily operations. This type of information is often more complex and varied than standard web-scraped content, making it a desirable asset for refining model performance.
This transaction reflects a broader trend where data is increasingly viewed as a key asset on a company’s balance sheet during liquidation. The competitive nature of the bid demonstrates that demand for this kind of information is rising; an artificial intelligence firm named Mercor had previously offered $7.5 million for the same data. The higher bid from Google suggests that established tech companies are willing to pay a premium to secure proprietary archives that can enhance their AI capabilities.
To address privacy and security concerns, the companies have stated that the data will be de-identified before use. This process is designed to strip out personally identifiable information, such as customer records or credit card details, to protect privacy. Despite these safeguards, the use of large-scale corporate data for AI training continues to face scrutiny. A potential risk for companies involved in such deals is the possibility of residual privacy leaks or the inadvertent inclusion of confidential business strategies or intellectual property that may remain in the archive even after the de-identification process.
The final approval of this sale will be a key event for the market to watch. If the court allows the transaction to proceed, it could set a precedent for how future bankruptcy cases handle the liquidation of digital assets. For investors, the development signals that the value of information held by distressed companies is changing, potentially creating a new market for specialized data archives in the technology sector.