AI Labs Acquire Data from Defunct Startups to Train Models
Current overview (not historical)
AI research labs are purchasing digital assets from bankrupt startups to feed their machine‑learning models. Companies such as SimpleClosure have processed more than 150 transactions, with an estimated $10 million to be returned by year‑end. The valuation of these data sets varies widely, often based on academic papers, and can reach hundreds of millions of dollars—Google reportedly paid $10 million for billions of emails and tool logs in the Spirit Airlines bankruptcy. While these deals provide a liquidation route for failed firms, they raise significant privacy and copyright concerns. Employees of the defunct companies frequently remain unaware of who ultimately owns or how their data will be used, creating a potential conflict between the interests of the original creators and the new data buyers.
Written with AI from the reporting · updated 10/14/2026
Expand the detailed fact route and sources
Story route
A station-by-station route from the first public evidence onward; every station carries its own evidence and uncertainty.
- 1Unclassified fact stationOct 13, 00:00recorded Oct 14, 22:23single source
AI labs buying data from dead startups to train models
Read original source ↗ MIT Technology Review · AIReplay this moment in the explorerEvidence & revisions · 1 source records
- Subject：
- SimpleClosure
- Action：
- brokered data sales of code and workspace data from dying startups
- Object：
- code base and workspace data (emails, chats, documents)
- Occurred (fact side)：
- Oct 13, 00:00
- Published (valid side)：
- Oct 13, 17:00
- First seen (known side)：
- Oct 14, 22:23
Fact stationSingle source · one-sided claimCopy may have been revised later; time replay is bounded. Layer labels group reading only, not causation between layers.
Uncertainty marks flag where to check the sources; they are not proven causation.
Report timeline
Explore every side through dated source reports.
- MIT Technology Review · AIRelated fact: single sourceAI Labs Purchase Data from Shutting Down Startups to Train Models
Report first recorded Oct 14, 22:23
Tech undertakers like SimpleClosure are selling digital archives, including code and workspace data from dying startups, to AI labs and mature tech companies. SimpleClosure has handled over 150 deals and expects to return over $10 million to startups by the end of 2026. Buyers, such as Surge AI, seek real-world data to build reinforcement-learning gyms, as AI training labs face limits with public and synthetic datasets. Valuations vary widely, as noted by Nick Heiner, because the performance gains from real-world data are not yet clearly tested.
Read original source ↗
Event discussion
Reader opinions are hosted on GitHub Discussions; posting requires a GitHub account. Comments are not verified facts.