micro1 Beats Google to Buy Spirit Records for $12.5M

micro1 has outbid Google with a $12.5 million offer to acquire the digital archives of defunct carrier Spirit Aviation, surpassing the tech giant’s previously agreed-upon $10 million bid. The AI training-data firm submitted its proposal in a recent court filing, aiming to secure the airline’s vast trove of internal communications while addressing concerns over data privacy and control that have plagued the earlier agreement.

Spirit Aviation Holdings ceased operations in May and is currently undergoing liquidation. The assets up for grabs are substantial, comprising approximately 500 million Microsoft Teams items, 100 million emails, and roughly 16 million customer chat sessions. While the monetary value of micro1’s bid is higher, the company’s strategy focuses heavily on restructuring how the data will be handled and who controls the process.

Under the terms of Google’s deal, Spirit was required to hand the material directly to parties designated by the buyer. Furthermore, Google selected and funded the third-party firm responsible for deidentifying the data, with those costs added on top of the purchase price rather than deducted from it. This arrangement drew criticism regarding transparency and oversight.

In contrast, micro1 proposes a more neutral governance structure. The new offer includes an ombudsman selected by Spirit’s own advisers, ensuring an independent party oversees the data handling process. Additionally, micro1 pledges to store the data within the United States, aiming to alleviate some jurisdictional anxieties.

The court filing also explicitly excludes sensitive categories such as disciplinary records, investigatory materials, and any content related to collective bargaining with unions. These exclusions address long-standing objections raised by labor groups, which had already challenged the Google sale on privacy grounds.

Google maintains that it will not receive any personal information from the dataset. A judge is scheduled to review the purchase on September 9. However, micro1 faces significant procedural hurdles. Courts rarely reopen auctions that have already concluded, meaning the challenger must overcome legal precedents rather than simply offering a better price.

Complicating matters further is the scope of the data itself. Google’s original agreement excluded customer chat sessions, loyalty records, and call recordings. Micro1’s offer specifically includes approximately 16 million chat sessions, creating a discrepancy in what is actually being sold despite the higher price tag.

The debate over these records highlights stark differences between American and European regulatory philosophies. In Europe, the concept of “deidentified” data is treated as a matter of capability rather than mere labeling. The Court of Justice ruled last year that pseudonymised data remains personal data if the recipient can realistically identify individuals. This makes the preservation of referential integrity, which keeps pseudonymous records linked across systems, a critical privacy risk under EU law.

Furthermore, the European Data Protection Board has stated that models trained on personal data are not automatically anonymous. Regulators may scrutinize whether training data was obtained lawfully, especially given the principle of purpose limitation. Records originally generated to fly aircraft and pay staff were not collected for AI training, meaning their reuse in the EU would require a separate legal basis.

None of these stringent European constraints apply to Spirit’s estate, which is wound up under American law. This legal distinction is why the situation plays out as a competitive bidding war in US courts rather than a regulatory compliance issue. If European laws governed this liquidation, the focus would likely shift entirely from price to lawful data processing capabilities.