<\/p>\n
SimpleClosure added data sales to its services in April. Its Asset Hub platform helps founders sell their code and workspace data\u2014emails, chats, documents. Buyers include foundation AI labs, vendors that construct simulated workspaces for AI training, and mature tech companies building their own models. Asset Hub has handled over 150 deals and expects to return over $10 million to startups by the end of 2026. CEO Dori Yona has seen these transactions, once rare and \u201csomewhat taboo,\u201d become increasingly routine, with 47% of SimpleClosure deals now including code or data sales.\u00a0<\/p>\n
And it\u2019s got company. Industry rival Sunset also has a rapidly growing data resale business. Founder and CEO Brendan Mahony says Sunset values a dying client\u2019s assets by head count, years in operation, volume of data, and how the team used its tools. Particle board, he says, is a good analogy: \u201cThis operational exhaust data is like the wood shavings that everyone thought had no value, but it actually has a ton.\u201d\u00a0<\/p>\n
Demand for defunct companies\u2019 data emerged suddenly last year, as AI training labs hit the limits of public and synthetic datasets. Surge AI, for example, buys data to build \u201creinforcement-learning gyms\u201d where clients\u2019 AI agents practice tasks\u2014think debugging a complex IT architecture or drafting appellate briefs. Nick Heiner, who leads the company\u2019s work on these environments, says he seeks data archives reflecting a range of situations, from disagreements to tricky judgment calls. Heiner explains that one company\u2019s Slack history, if it captures conversations with variety and depth, can be worth more than another\u2019s Slack, email, and calendars combined.<\/p>\n
Real archives avoid common errors in synthetic worlds, such as references to dates that don\u2019t exist, but the extent of performance gains remains unclear. No company has yet released tests comparing results of real-world data training with the alternatives. Pricing reflects this uncertainty: Buyers quote wildly different numbers, Heiner says, because valuations are based on \u201cresearch theses\u201d estimating how many useful tasks come from, say, 100,000 lines of code.<\/p>\n
That hasn\u2019t stopped bigger deals from happening. This past August, as part of Spirit Airlines\u2019 bankruptcy, Google outbid the AI data company Mercor to buy the carrier\u2019s decades of data, including 100 million emails, 500 million Teams messages, and 176,000 employee records. The price tag: $10 million. And Bloomberg<\/em> reported in September that SpaceX\u2019s AI unit is weighing buying data from struggling startups.\u00a0<\/p>\n Legally, this all could get iffy. Spirit\u2019s flight attendants\u2019 union objected to the sale on privacy grounds, and Greg Ewing, an attorney who advises AI startups, has called such data \u201clegally radioactive.\u201d Suppliers employ data scientists to scrub personally identifiable information and exclude HR records, and buyers say they scrub too. But Ewing points out some of the private details that lurk in emails and calendars: doctor appointments, school pickup arrangements, date-night plans. He describes a \u201cWild West of AI training\u201d that is only starting to give way to proper licensing agreements and vetted use.<\/p>\n There\u2019s also the question of who profits from these windfalls. The proceeds typically flow to senior creditors and investors\u2014not employees. To Ifeoma Ajunwa, a law professor at Emory University and author of The<\/em> Quantified Worker<\/em>, this represents a form of \u201ccaptured capital.\u201d Legal precedent allows for monitoring to help workers do their jobs, but once a firm is out of business, that justification vanishes. \u201cEmployers should have express consent before selling workers\u2019 data,\u201d says Ajunwa, but she adds that there\u2019s no federal law banning such sales.<\/p>\n While many see this new market as a chance for failing businesses to recover some value, these wind-down sales also have limitations. Founders have little to no visibility into who is buying the data and what they plan to do with it. Plus, there\u2019s little assurance they\u2019re getting a fair offer. Data prices will likely stabilize in the coming year, though, unless an AI bubble pops first. Either way, tech\u2019s new undertakers will be there to appraise what\u2019s left.\u00a0<\/p>\n Matthew King is a technology and environmental journalist based in Baltimore.<\/em>