Bankrupt Firm Data AI Acquisition
AI companies are increasingly competing for access to bankrupt firms' internal data as a scarce and valuable resource for model training
Too little corroboration in the last 3 days to call a trend (7 articles). Watching for it to gain traction.
Mainstream financial press is carrying this — attention has broadened beyond specialist outlets.
"AI firms are increasingly gobbling up defunct businesses' internal communications— Slack archives, Jira tickets, email threads, Google Drive files and more—to improve their models' skillsets... The competing bid from Mercor shows how AI companies are competing for this valuable new source of training material."
"Google is paying $10 million for Spirit Airlines' business data, including internal communications and company records that it plans to use to improve its AI models. We acquired part of an enterprise dataset from Spirit Airlines, which can be helpful in improving our products and AI models."
"The sale comes as AI companies increasingly pay for access to data used to train and improve their models. In February 2024, Reddit gave Google access to nearly two decades of user-generated content in a deal reportedly worth $60 million a year."
"If proprietary workplace data proves useful for training AI agents, failed companies may possess something AI developers increasingly want. Their emails, chats, spreadsheets, workflows, and operational histories could become training material unavailable anywhere on the public web."
"Google is buying all of Spirit Airlines' data for $10 million to continue training its AI models following the halting of its operations in May. A Google spokesperson told CNN: 'We acquired part of an enterprise dataset from Spirit Airlines, which can be helpful in improving our products and AI models.'"
"That is increasingly important as technology companies search for additional sources of training materials beyond publicly available internet content. Real-world corporate records can reveal relationships between decisions, processes and outcomes that are difficult to reproduce using publicly available text alone."
"Alphabet's Google agreed to pay $10 million for a vast trove of de-identified internal data from bankrupt Spirit Airlines, including roughly 100 million employee emails, 500 million Microsoft Teams chats, spreadsheets, calendars and operational records. The company plans to use the material for product development and AI model training."