1. What happened
Bankruptcy asset auction. Google was selected at $10 million; Mercor was alternate at $7.5 million. micro1 later noticed a $12.5 million competing bid. No sale order had been entered as of the review date.
2. Buyer / seller
Google LLC (selected bidder; challenged) is the AI buyer or model-development side. Spirit Aviation Holdings / Spirit Airlines owns or contributes the proprietary corpus.
3. Economics
Google selected bid; court approval pending. Later micro1 notice proposed $12.5 million. Economics scope: dataset and related internal software specific. Strategic Signal has not estimated undisclosed consideration.
4. Proprietary data
A deidentified archive spanning roughly 34 years, reported to include about 100 million emails, 500 million Teams messages, 30 million lines of source code, operational data, documents, and internally developed software. Customer databases are outside scope. Historical depth: Approximately 34 years. Refreshability: none airline ceased operations.
5. Rights and restrictions
Independent third-party deidentification contemplated before delivery; final restrictions remain subject to court approval. Ownership: yes if sale closes. Training: yes explicit. Post-training: yes general model improvement.
6. Why the data matters for AI
Train and improve AI systems and productivity products on real enterprise work, software, communication, and operational sequences. High: communications, code, operations, scheduling, finance, and maintenance-adjacent workflows create a dense cross-functional operating history.
7. Asset Score — Pending/100
Legacy score withdrawn pending evidence-linked scoring and human approval. The superseded v1.0.0 release retains the original values. Score coverage: 0%.
| Factor | Weight | Rating / 5 | Points |
|---|---|---|---|
| Uniqueness / scarcity | 12 | Unknown | — |
| Historical depth | 8 | Unknown | — |
| Decision → outcome richness | 13 | Unknown | — |
| Decision Value Density | 13 | Unknown | — |
| Real-world grounding | 10 | Unknown | — |
| Domain economic value | 9 | Unknown | — |
| Refreshability | 8 | Unknown | — |
| Proprietary advantage | 9 | Unknown | — |
| Model-learning usefulness | 10 | Unknown | — |
| Non-replicability | 4 | Unknown | — |
| Rights usability | 4 | Unknown | — |
8. Transaction Signal Score — Pending/100
Legacy score withdrawn pending evidence-linked scoring and human approval. The superseded v1.0.0 release retains the original values. Score coverage: 0%.
| Factor | Weight | Rating / 5 | Points |
|---|---|---|---|
| Disclosed economics | 16 | Unknown | — |
| Clean price discovery | 16 | Unknown | — |
| Separability of data consideration | 15 | Unknown | — |
| Explicit training / model-improvement use | 14 | Unknown | — |
| Clarity of rights purchased | 12 | Unknown | — |
| Competing bids | 10 | Unknown | — |
| Strategic buyer quality | 6 | Unknown | — |
| Repeatability / precedent value | 6 | Unknown | — |
| Independent model-uplift evidence | 5 | Unknown | — |
9. M&A / valuation implications
Provides unusually clean evidence that a defunct company's data and institutional memory can be auctioned separately from its operating business, while exposing privacy and chain-of-title diligence as value-critical.
10. Evidence boundary
Bid amounts, selected/alternate bidders, asset scope, and pending status are sourced facts. Model-use and valuation implications beyond disclosed purpose are Strategic Signal analysis.