AI TrackerFrontier desk · AEST
Back to news
normalaicryptotech

AI labs are buying data from dead startups to train their models

MIT Technology Review · 13 Oct 2026, 19:00 AEST

AI labs are increasingly purchasing data from defunct startups to enhance their models. This trend has emerged as a new sector in tech, with companies facilitating the sale of digital archives from these businesses.

Key facts

Read aloud

A I labs are increasingly purchasing data from defunct startups to enhance their models. This trend has emerged as a new sector in tech, with companies facilitating the sale of digital archives from these businesses. When Keelan Miskell began winding down his crypto startup this year, he didn’t think its data was worth much: the users were anonymous, the company just a few years old. Estimates from Carta and AngelList show shutdowns rising between twenty six percent and fifty six percent annually in recent years, and an end-of-life ecosystem—including coaches, 'death doulas,' and firms that help handle the legal, tax, and operational headaches of going belly-up—has emerged. SimpleClosure added data sales to its services in April. Its Asset Hub platform helps founders sell their code and workspace data—emails, chats, documents. Demand for defunct companies’ data emerged suddenly last year, as A I training labs hit the limits of public and synthetic datasets. Legally, this all could get iffy. Spirit’s flight attendants’ union objected to the sale on privacy grounds, and Greg Ewing, an attorney who advises A I startups, has called such data 'legally radioactive.'

Sources in this cluster