Founders winding down companies assumed their records held little value, and AI labs see training material instead.
AI labs are buying data from startups that have shut down, treating archives that founders once considered near-worthless as training material for new models. Keelan Miskell began winding down his crypto startup and did not expect its data to hold much value, yet that record now has buyers.
The story comes from MIT Technology Review’s The Download, a weekday newsletter that covers the technology industry. The same edition also covers interspecies communication, which places the data-buying trend alongside research on animal signals.
For builders, the practical lesson is that a shutdown plan should account for data assets. Teams that kept clean records of how they collected and stored information hold something labs want, and operators should price that material before the servers go dark.
Watch whether labs formalize channels for acquiring data from closing companies, since a founder’s final quarter is when those assets change hands. More wind-downs will test whether buyers keep paying for archives that look unremarkable at first glance.
What matters
- AI labs are buying data held by startups that have shut down or wound down operations.
- Operators sitting on old company data should treat those archives as a sellable asset during wind-down.
- Watch whether more labs disclose where they source training data as startup archives dry up.
Why it matters
Watch whether more labs disclose where they source training data as startup archives dry up.
This GenAI News article was prepared in original wording using reporting and materials published by MIT Technology Review AI. Source reference: https://www.technologyreview.com/2026/10/28/1147767/the-download-dead-startups-fueling-ai-interspecies-communication/.
Drafted by the GenAI News review pipeline.