{ "name": "Programs that license company data for AI training", "publisher": "Marin Data", "updated": "2026-11-02", "license": "CC BY 4.0", "page": "https://marindata.ai/buyers", "readme": "https://marindata.ai/data/README.md", "note": "Each term is what the program publishes. \"Not published\" means we found no public figure. Marin Data is not an agent of any program listed.", "programs": [ { "slug": "aitrainer-work", "name": "aitrainer.work", "kind": "Intermediary", "url": "https://marindata.ai/buyers/aitrainer-work", "summary": "aitrainer.work Enterprise is a free matching form credited to Pietro Romeo that shows which data buyers fit a company. It offers an introduction to Turing on opt-in and plain application links for most other buyers.", "goodFitFor": "Owners who want a short list of programs that might fit before they speak to any buyer.", "watchFor": [ "The page does not say whether aitrainer.work is paid by the buyers it lists (observed Oct 16, 2026).", "Only Turing gets an introduction; other matches are links to apply yourself (aitrainer.work, observed Oct 16, 2026)." ], "lastVerified": "2026-11-02", "termsPublished": 3, "notes": [], "terms": { "whatTheyBuy": { "label": "What they license", "value": "Does not buy. It matches companies holding repos, datasets, tasksets, closed-company records or live workflows to buyers", "source": { "label": "aitrainer.work, enterprise matching", "href": "https://aitrainer.work/enterprise/refer", "date": "2026-10-16" } }, "priceOrBands": { "label": "Price or bands", "value": "Free for companies", "source": { "label": "aitrainer.work, enterprise matching", "href": "https://aitrainer.work/enterprise/refer", "date": "2026-10-16" } }, "headcountFloor": { "label": "Headcount floor", "value": "Not published", "source": null }, "historyFloor": { "label": "Years of history", "value": "Not published", "source": null }, "geography": { "label": "Geography", "value": "Not published", "source": null }, "exclusivity": { "label": "Exclusivity", "value": "Not published", "source": null }, "paymentTiming": { "label": "Payment timing", "value": "Not published", "source": null }, "whoDeidentifies": { "label": "Who de-identifies", "value": "The buyer; the form asks about the company and never for its data", "source": { "label": "aitrainer.work, enterprise matching", "href": "https://aitrainer.work/enterprise/refer", "date": "2026-10-16" } }, "referralBounty": { "label": "Referral bounty", "value": "Not published", "source": null }, "publishedTermsPage": { "label": "Published terms page", "value": "Not published", "source": null } }, "sources": [ { "label": "aitrainer.work, enterprise matching", "href": "https://aitrainer.work/enterprise/refer", "date": "2026-10-16" }, { "label": "aitrainer.work, how companies get paid for operational data", "href": "https://aitrainer.work/enterprise/guides/how-companies-get-paid-for-operational-data", "date": "2026-10-04" }, { "label": "aitrainer.work, AI labs paying for enterprise workflow data", "href": "https://aitrainer.work/news/ai-labs-paying-enterprise-workflow-data-2026", "date": "2026-06-29" } ] }, { "slug": "anthroly", "name": "Anthroly", "kind": "Data partner", "url": "https://marindata.ai/buyers/anthroly", "summary": "Anthroly licenses operational records from US companies and supplies them to AI labs. It prices each set of records individually after reading a de-identified sample.", "goodFitFor": "US companies with records across many systems that want a price set on their own records, not a band.", "watchFor": [ "No price band, headcount floor or history floor is published (Anthroly, observed Oct 19, 2026).", "The site names no legal entity behind Anthroly (Anthroly, observed Oct 19, 2026)." ], "lastVerified": "2026-11-02", "termsPublished": 7, "notes": [ { "text": "To start, a company sends its name, the systems it uses and roughly how far back the records go. Anthroly asks for no files at that stage.", "source": { "label": "Anthroly, homepage", "href": "https://www.anthroly.com/", "date": "2026-10-19" } } ], "terms": { "whatTheyBuy": { "label": "What they license", "value": "Team chat, email, support tickets, CRM history, calls and meetings, code and engineering history, documents and wikis, finance and ERP exports, field records and design files", "source": { "label": "Anthroly, homepage", "href": "https://www.anthroly.com/", "date": "2026-10-19" } }, "priceOrBands": { "label": "Price or bands", "value": "No band published; each record set is priced after a de-identified sample is read", "source": { "label": "Anthroly, homepage", "href": "https://www.anthroly.com/", "date": "2026-10-19" } }, "headcountFloor": { "label": "Headcount floor", "value": "Not published", "source": null }, "historyFloor": { "label": "Years of history", "value": "Not published", "source": null }, "geography": { "label": "Geography", "value": "Companies headquartered and operating in the US", "source": { "label": "Anthroly, homepage", "href": "https://www.anthroly.com/", "date": "2026-10-19" } }, "exclusivity": { "label": "Exclusivity", "value": "Licensed, not sold, for a defined material, term and use", "source": { "label": "Anthroly, homepage", "href": "https://www.anthroly.com/", "date": "2026-10-19" } }, "paymentTiming": { "label": "Payment timing", "value": "After delivery and approval", "source": { "label": "Anthroly, homepage", "href": "https://www.anthroly.com/", "date": "2026-10-19" } }, "whoDeidentifies": { "label": "Who de-identifies", "value": "Identifiers are removed before any data moves; the page does not say who does it", "source": { "label": "Anthroly, homepage", "href": "https://www.anthroly.com/", "date": "2026-10-19" } }, "referralBounty": { "label": "Referral bounty", "value": "Not published", "source": null }, "publishedTermsPage": { "label": "Published terms page", "value": "Terms linked in the site footer", "source": { "label": "Anthroly, homepage", "href": "https://www.anthroly.com/", "date": "2026-10-19" } } }, "sources": [ { "label": "Anthroly, homepage", "href": "https://www.anthroly.com/", "date": "2026-10-19" } ] }, { "slug": "appen", "name": "Appen", "kind": "Data partner", "url": "https://marindata.ai/buyers/appen", "summary": "Appen is an ASX-listed AI training data company founded in 1996. Its Enterprise Data program licenses operational documentation, decision and escalation records and systems of record to AI labs, and Appen pays for extraction, anonymization and transfer.", "goodFitFor": "Companies, including those winding down, whose value sits in documented procedures and in approval, exception and escalation trails.", "watchFor": [ "Appen also sells enterprise company data to AI labs (Who Sells Data, Oct 14, 2026).", "Appen's page names no AI labs and no prices; both are set per deal (Appen, observed Oct 17, 2026)." ], "lastVerified": "2026-11-02", "termsPublished": 3, "notes": [ { "text": "Appen groups what it licenses by record type. Operational documentation covers procedures, knowledge bases, templates, training material and QA steps. Decision and escalation records come from support desks, ITSM, CRM, project trackers and messaging. Systems of record include reconciliations, executed agreements, reporting lineage and HR, payroll and procurement data.", "source": { "label": "Appen, enterprise data", "href": "https://www.appen.com/enterprise-data", "date": "2026-10-17" } }, { "text": "The process: register interest, hear back from the partnerships team within two business days, sign a mutual NDA and written terms, then set the scope by system, team, workflow and time period. Appen extracts through read-only, encrypted pipelines with per-engagement credentials. The seller reviews a sample pack, and Appen says nothing is used without that sign-off.", "source": { "label": "Appen, enterprise data", "href": "https://www.appen.com/enterprise-data", "date": "2026-10-17" } }, { "text": "Retention periods and deletion terms are agreed in writing before extraction starts. The page gives no default period, so the seller has to ask for one.", "source": { "label": "Appen, enterprise data", "href": "https://www.appen.com/enterprise-data", "date": "2026-10-17" } }, { "text": "Appen says the seller keeps ownership and grants \"a defined licence for a defined use\". The page does not say whether that licence is exclusive.", "source": { "label": "Appen, enterprise data", "href": "https://www.appen.com/enterprise-data", "date": "2026-10-17" } }, { "text": "A company winding down gets a faster path and can often license its full operational history instead of a scoped subset. Appen notes that records often become unrecoverable once systems are shut off and software licences lapse.", "source": { "label": "Appen, enterprise data", "href": "https://www.appen.com/enterprise-data", "date": "2026-10-17" } }, { "text": "Appen lists ISO 27001 and SOC 2 Type II certification and a HIPAA-compliant environment.", "source": { "label": "Appen, data security", "href": "https://www.appen.com/data-security", "date": "2026-10-17" } }, { "text": "In its H1 FY26 results Appen reported revenue of $119.9M. Appen China grew 80% to $76.2M, Appen Global fell 27% to $43.7M, and the filing credits \"our relationship with Chinese model builders\". The filing does not mention the Enterprise Data program. A seller who cares where its records end up should ask which AI labs, and which countries, are in scope.", "source": { "label": "Appen, H1 FY26 results (ASX announcement)", "href": "https://announcements.asx.com.au/asxpdf/20260827/pdf/0739d4q7lyh3w2.pdf", "date": "2026-08-27" } } ], "terms": { "whatTheyBuy": { "label": "What they license", "value": "Operational documentation, decision and escalation records, and systems of record", "source": { "label": "Appen, enterprise data", "href": "https://www.appen.com/enterprise-data", "date": "2026-10-17" } }, "priceOrBands": { "label": "Price or bands", "value": "Not published", "source": null }, "headcountFloor": { "label": "Headcount floor", "value": "Not published", "source": null }, "historyFloor": { "label": "Years of history", "value": "Not published", "source": null }, "geography": { "label": "Geography", "value": "Not published", "source": null }, "exclusivity": { "label": "Exclusivity", "value": "Not published", "source": null }, "paymentTiming": { "label": "Payment timing", "value": "On acceptance, under terms agreed in writing before extraction begins", "source": { "label": "Appen, enterprise data", "href": "https://www.appen.com/enterprise-data", "date": "2026-10-17" } }, "whoDeidentifies": { "label": "Who de-identifies", "value": "Appen detects and removes personal data, with synthetic rewriting where the workflow structure has to survive redaction", "source": { "label": "Appen, enterprise data", "href": "https://www.appen.com/enterprise-data", "date": "2026-10-17" } }, "referralBounty": { "label": "Referral bounty", "value": "Not published", "source": null }, "publishedTermsPage": { "label": "Published terms page", "value": "Not published", "source": null } }, "sources": [ { "label": "Appen, enterprise data", "href": "https://www.appen.com/enterprise-data", "date": "2026-10-17" }, { "label": "Appen, data security", "href": "https://www.appen.com/data-security", "date": "2026-10-17" }, { "label": "Appen, H1 FY26 results (ASX announcement)", "href": "https://announcements.asx.com.au/asxpdf/20260827/pdf/0739d4q7lyh3w2.pdf", "date": "2026-08-27" }, { "label": "Who Sells Data, buying programs register", "href": "https://whosellsdata.com/buying", "date": "2026-10-14" } ] }, { "slug": "centific", "name": "Centific", "kind": "Data partner", "url": "https://marindata.ai/buyers/centific", "summary": "Centific is a privately held AI data services company based in the Seattle area. It organizes its enterprise data program by industry, anonymizes the operational records it collects and licenses them to frontier labs, paying on acceptance.", "goodFitFor": "Enterprises in finance, healthcare operations, sales, people operations or IT service management that want a recurring refresh, including those winding down.", "watchFor": [ "Its healthcare area includes claims and de-identified clinical documentation (Who Sells Data, Oct 14, 2026).", "The program page states no retention or deletion terms (Centific, observed Oct 17, 2026)." ], "lastVerified": "2026-11-02", "termsPublished": 3, "notes": [ { "text": "Centific sorts its program by industry. Finance and accounting covers approval and exception chains, reconciliations, redline history, close-to-report, procurement and AP/AR. Healthcare operations covers claims, revenue-cycle work, care coordination and de-identified clinical documentation. IT service management includes incidents, escalations, reopened tickets and failed changes.", "source": { "label": "Centific, enterprise data", "href": "https://www.centific.com/enterprise-data", "date": "2026-10-17" } }, { "text": "After a first call, Centific gives a written estimate of the dataset's value. It names five price factors: volume, the number of connected applications, work complexity, data quality and \"availability of comparable data\".", "source": { "label": "Centific, enterprise data", "href": "https://www.centific.com/enterprise-data", "date": "2026-10-17" } }, { "text": "It connects read-only to what it says are 188 supported enterprise applications. Exclusions are enforced in the extraction code, and the seller gets an audit trail of every object read.", "source": { "label": "Centific, enterprise data", "href": "https://www.centific.com/enterprise-data", "date": "2026-10-17" } }, { "text": "Centific says most partnerships reach a first accepted contribution in six to ten weeks, and most partners then move to a quarterly refresh.", "source": { "label": "Centific, enterprise data", "href": "https://www.centific.com/enterprise-data", "date": "2026-10-17" } }, { "text": "Employees can be paid to annotate their own anonymized records. That work is optional and priced separately from the data license.", "source": { "label": "Centific, enterprise data", "href": "https://www.centific.com/enterprise-data", "date": "2026-10-17" } }, { "text": "Companies winding down are offered a faster path under an asset-purchase agreement. Elsewhere the page describes a license, so a seller should check whether that agreement transfers ownership of the records.", "source": { "label": "Centific, enterprise data", "href": "https://www.centific.com/enterprise-data", "date": "2026-10-17" } }, { "text": "Centific raised a $60M Series A led by Granite Asia in June 2025.", "source": { "label": "Centific closes $60M Series A (release via AAP)", "href": "https://www.aap.com.au/aapreleases/cision20250625ae17819", "date": "2025-06-25" } }, { "text": "On Aug 3, 2026 it announced that Chuen Hong Lew will succeed Venkat Rangapuram as CEO.", "source": { "label": "Centific, news and press (CEO succession, Aug 3, 2026)", "href": "https://www.centific.com/news", "date": "2026-10-17" } }, { "text": "Its locations page gives a Redmond, Washington address, with offices in India, mainland China, Taiwan, Spain, Malaysia, Serbia and Singapore.", "source": { "label": "Centific, locations", "href": "https://www.centific.com/company/locations", "date": "2026-10-17" } } ], "terms": { "whatTheyBuy": { "label": "What they license", "value": "Operational data in finance and accounting, healthcare operations, sales and deal workflows, people operations, and IT service management", "source": { "label": "Centific, enterprise data", "href": "https://www.centific.com/enterprise-data", "date": "2026-10-17" } }, "priceOrBands": { "label": "Price or bands", "value": "Not published", "source": null }, "headcountFloor": { "label": "Headcount floor", "value": "Not published", "source": null }, "historyFloor": { "label": "Years of history", "value": "Not published", "source": null }, "geography": { "label": "Geography", "value": "Not published", "source": null }, "exclusivity": { "label": "Exclusivity", "value": "Not published", "source": null }, "paymentTiming": { "label": "Payment timing", "value": "On acceptance of the anonymized data, after the seller approves a sample pack", "source": { "label": "Centific, enterprise data", "href": "https://www.centific.com/enterprise-data", "date": "2026-10-17" } }, "whoDeidentifies": { "label": "Who de-identifies", "value": "Centific anonymizes the data automatically, with synthetic rewriting where needed", "source": { "label": "Centific, enterprise data", "href": "https://www.centific.com/enterprise-data", "date": "2026-10-17" } }, "referralBounty": { "label": "Referral bounty", "value": "Not published", "source": null }, "publishedTermsPage": { "label": "Published terms page", "value": "Not published", "source": null } }, "sources": [ { "label": "Centific, enterprise data", "href": "https://www.centific.com/enterprise-data", "date": "2026-10-17" }, { "label": "Centific closes $60M Series A (release via AAP)", "href": "https://www.aap.com.au/aapreleases/cision20250625ae17819", "date": "2025-06-25" }, { "label": "Centific, news and press (CEO succession, Aug 3, 2026)", "href": "https://www.centific.com/news", "date": "2026-10-17" }, { "label": "Centific, locations", "href": "https://www.centific.com/company/locations", "date": "2026-10-17" }, { "label": "Who Sells Data, buying programs register", "href": "https://whosellsdata.com/buying", "date": "2026-10-14" } ] }, { "slug": "collinear", "name": "Collinear AI", "kind": "Data partner", "url": "https://marindata.ai/buyers/collinear", "summary": "Collinear AI licenses code repositories and business app history to build AI training and evaluation environments. It quotes an initial range within a day and says signing takes about a week from the first call.", "goodFitFor": "Software companies with human-written code and linked tickets and pull requests, including teams below the 20-person floors of other programs.", "watchFor": [ "Personal data is removed after export, so raw data leaves the company first (Collinear, observed Oct 16, 2026).", "AI-generated code and loose documents without company context are not wanted (Collinear, observed Oct 16, 2026)." ], "lastVerified": "2026-11-02", "termsPublished": 5, "notes": [], "terms": { "whatTheyBuy": { "label": "What they license", "value": "Codebases with pull requests, linked tickets, CI logs and coding-agent sessions, plus history from apps such as Slack, Gmail, Drive, Notion, HubSpot, Salesforce, Linear and Jira", "source": { "label": "Collinear AI, data acquisition", "href": "https://collinear.ai/data-acquisition", "date": "2026-10-16" } }, "priceOrBands": { "label": "Price or bands", "value": "Not published", "source": null }, "headcountFloor": { "label": "Headcount floor", "value": "No floor; prices rise with 15+ people at peak", "source": { "label": "Collinear AI, data acquisition", "href": "https://collinear.ai/data-acquisition", "date": "2026-10-16" } }, "historyFloor": { "label": "Years of history", "value": "No floor; prices rise with several years of history", "source": { "label": "Collinear AI, data acquisition", "href": "https://collinear.ai/data-acquisition", "date": "2026-10-16" } }, "geography": { "label": "Geography", "value": "Not published", "source": null }, "exclusivity": { "label": "Exclusivity", "value": "Not published", "source": null }, "paymentTiming": { "label": "Payment timing", "value": "After Collinear confirms delivery, as set in the contract", "source": { "label": "Collinear AI, data acquisition", "href": "https://collinear.ai/data-acquisition", "date": "2026-10-16" } }, "whoDeidentifies": { "label": "Who de-identifies", "value": "Collinear, after export; names become consistent pseudonyms", "source": { "label": "Collinear AI, data acquisition", "href": "https://collinear.ai/data-acquisition", "date": "2026-10-16" } }, "referralBounty": { "label": "Referral bounty", "value": "Not published", "source": null }, "publishedTermsPage": { "label": "Published terms page", "value": "Not published", "source": null } }, "sources": [ { "label": "Collinear AI, data acquisition", "href": "https://collinear.ai/data-acquisition", "date": "2026-10-16" }, { "label": "Who Sells Data, buying programs register", "href": "https://whosellsdata.com/buying", "date": "2026-10-14" } ] }, { "slug": "data-licensing-report", "name": "Data Licensing Report (sellcompanydatatoai.io)", "kind": "Intermediary", "url": "https://marindata.ai/buyers/data-licensing-report", "summary": "Data Licensing Report compares companies that license business records for AI and takes applications through a three-field form. It tracks 25 providers with 609 cited sources and is free for businesses.", "goodFitFor": "Owners who want to compare providers before they apply anywhere.", "watchFor": [ "Billy Liu, named on the site, earns commission on some Telegraph Lab deals; no rate is given (Data Licensing Report, updated Oct 14, 2026)." ], "lastVerified": "2026-11-02", "termsPublished": 3, "notes": [], "terms": { "whatTheyBuy": { "label": "What they license", "value": "Does not buy. It reviews company profiles and then discusses licensing routes", "source": { "label": "Data Licensing Report, homepage", "href": "https://sellcompanydatatoai.io", "date": "2026-10-16" } }, "priceOrBands": { "label": "Price or bands", "value": "Free for businesses; its fee comes from a completed deal", "source": { "label": "Data Licensing Report, how we make money", "href": "https://sellcompanydatatoai.io/how-we-make-money/", "date": "2026-10-14" } }, "headcountFloor": { "label": "Headcount floor", "value": "Not published", "source": null }, "historyFloor": { "label": "Years of history", "value": "Not published", "source": null }, "geography": { "label": "Geography", "value": "Not published", "source": null }, "exclusivity": { "label": "Exclusivity", "value": "Not published", "source": null }, "paymentTiming": { "label": "Payment timing", "value": "Not published", "source": null }, "whoDeidentifies": { "label": "Who de-identifies", "value": "Not published", "source": null }, "referralBounty": { "label": "Referral bounty", "value": "Not published", "source": null }, "publishedTermsPage": { "label": "Published terms page", "value": "\"How we make money\" page, updated Oct 14, 2026", "source": { "label": "Data Licensing Report, how we make money", "href": "https://sellcompanydatatoai.io/how-we-make-money/", "date": "2026-10-14" } } }, "sources": [ { "label": "Data Licensing Report, homepage", "href": "https://sellcompanydatatoai.io", "date": "2026-10-16" }, { "label": "Data Licensing Report, how we make money", "href": "https://sellcompanydatatoai.io/how-we-make-money/", "date": "2026-10-14" } ] }, { "slug": "datapayd", "name": "Datapayd", "kind": "Intermediary", "url": "https://marindata.ai/buyers/datapayd", "summary": "Datapayd is a single-page intake form, registered in September 2026, that promises a private offer from unnamed AI buyers. It asks for company basics and a per-tool data inventory and says it replies within five business days.", "goodFitFor": "Owners comfortable sharing a data inventory before any buyer, fee or term is disclosed.", "watchFor": [ "The page names no legal entity, address or buyer, and states no fee beyond \"no upfront cost\" (Datapayd, observed Oct 17, 2026).", "Its /terms, /privacy and /about pages returned 404 (Datapayd, observed Oct 17, 2026)." ], "lastVerified": "2026-11-02", "termsPublished": 1, "notes": [ { "text": "The datapayd.com domain was registered on Sep 24, 2026, with the registrant hidden behind a privacy service.", "source": { "label": "Verisign RDAP, datapayd.com registration record", "href": "https://rdap.verisign.com/com/v1/domain/datapayd.com", "date": "2026-10-17" } }, { "text": "Step one of the form asks for company name, year founded, approximate headcount, industry and country. Step two is an inventory per application: years of data, size in GB, TB or records, what the tool is used for and how soon it could be exported. The page says no files are uploaded at this stage.", "source": { "label": "Datapayd, homepage", "href": "https://datapayd.com", "date": "2026-10-17" } }, { "text": "The tools it asks about include Google Drive, SharePoint, Slack, Teams, Gmail, Outlook, Salesforce, HubSpot, NetSuite, SAP, QuickBooks, Klaviyo, Notion, Confluence, Zendesk and Intercom. It also lists EMR and EHR systems such as Epic and athenahealth, and says it will walk through HIPAA requirements before anything is shared.", "source": { "label": "Datapayd, homepage", "href": "https://datapayd.com", "date": "2026-10-17" } }, { "text": "The page says \"Bigger and older is better\" and that you \"approve every term\". It does not say whether its own fee comes out of the buyer's offer.", "source": { "label": "Datapayd, homepage", "href": "https://datapayd.com", "date": "2026-10-17" } }, { "text": "Datapayd is not on the Who Sells Data register, and we found no press coverage or other directory listing for it.", "source": { "label": "Who Sells Data, buying programs register (updated Oct 15, 2026)", "href": "https://whosellsdata.com/buying", "date": "2026-10-17" } } ], "terms": { "whatTheyBuy": { "label": "What they license", "value": "Business data across the applications a company uses, from files, chat and email to CRM, ERP, support tools and healthcare EMR/EHR systems; industries listed include healthcare, manufacturing, logistics, accounting, legal, finance, insurance and construction", "source": { "label": "Datapayd, homepage", "href": "https://datapayd.com", "date": "2026-10-17" } }, "priceOrBands": { "label": "Price or bands", "value": "Not published", "source": null }, "headcountFloor": { "label": "Headcount floor", "value": "Not published", "source": null }, "historyFloor": { "label": "Years of history", "value": "Not published", "source": null }, "geography": { "label": "Geography", "value": "Not published", "source": null }, "exclusivity": { "label": "Exclusivity", "value": "Not published", "source": null }, "paymentTiming": { "label": "Payment timing", "value": "Not published", "source": null }, "whoDeidentifies": { "label": "Who de-identifies", "value": "Not published", "source": null }, "referralBounty": { "label": "Referral bounty", "value": "Not published", "source": null }, "publishedTermsPage": { "label": "Published terms page", "value": "Not published", "source": null } }, "sources": [ { "label": "Datapayd, homepage", "href": "https://datapayd.com", "date": "2026-10-17" }, { "label": "Verisign RDAP, datapayd.com registration record", "href": "https://rdap.verisign.com/com/v1/domain/datapayd.com", "date": "2026-10-17" }, { "label": "Who Sells Data, buying programs register (updated Oct 15, 2026)", "href": "https://whosellsdata.com/buying", "date": "2026-10-17" } ] }, { "slug": "defined-ai", "name": "Defined.ai", "kind": "Marketplace", "url": "https://marindata.ai/buyers/defined-ai", "summary": "Defined.ai, the Seattle company formerly called DefinedCrowd, runs a marketplace that sells partners' audio, video, image, text and multimodal data to AI developers and pays licensing fees from those sales. It lists cinematic, anime, coding and STEM data as in high demand.", "goodFitFor": "Businesses with media, code or STEM datasets they have the consent and rights to license for AI training.", "watchFor": [ "The $600K+ average partner revenue on its page is Defined.ai's own figure (Who Sells Data, Oct 14, 2026).", "No revenue share, payment timing or seller agreement is published (Defined.ai, observed Oct 17, 2026)." ], "lastVerified": "2026-11-02", "termsPublished": 2, "notes": [ { "text": "Applications go through a Jira service-desk account. Defined.ai then researches commercial demand, evaluates the data's quality, volume, organization and metadata, and checks licensing rights, legal, compliance and cybersecurity.", "source": { "label": "Defined.ai, partnership programs", "href": "https://defined.ai/partnership-programs", "date": "2026-10-17" } }, { "text": "Commercial terms depend on dataset type, usage and customer needs, and Defined.ai says the partner approves them. It says it can consider usage restrictions case by case. There is no fixed minimum dataset size.", "source": { "label": "Defined.ai, partnership programs", "href": "https://defined.ai/partnership-programs", "date": "2026-10-17" } }, { "text": "All data must have proper consent, documentation and the right to commercialize it for AI training. Applicants with cinematic, anime, coding or STEM data get an expedited onboarding path.", "source": { "label": "Defined.ai, partnership programs", "href": "https://defined.ai/partnership-programs", "date": "2026-10-17" } }, { "text": "The published Data License Agreement is the buyer-side contract. It gives buyers a non-exclusive license and lets them commercially use models trained on the data, and either side can end it on 60 days' notice. A seller's records can therefore stay inside a buyer's models after a license ends.", "source": { "label": "Defined.ai, data license agreement (buyer side, effective Nov 30, 2022)", "href": "https://defined.ai/data-license-agreement", "date": "2026-10-17" } }, { "text": "Reuters reported in April 2024 that Defined.ai licenses data to Google, Meta, Apple, Amazon and Microsoft, at about $1 to $2 per image, $2 to $4 per short video and $0.001 per word of text. Those are buyer prices; the seller's share was not reported.", "source": { "label": "The Decoder, on Reuters report of AI training data prices", "href": "https://the-decoder.com/tech-giants-are-on-a-billion-dollar-shopping-spree-for-ai-training-data/", "date": "2024-04-06" } }, { "text": "In January 2026 Defined.ai reported 65% revenue growth, a 1,200% rise in third-party partner datasets and several partners earning more than $1 million a year. It gave no total revenue figure.", "source": { "label": "Defined.ai, reports 65% revenue growth", "href": "https://defined.ai/press-room/defined-ai-reports-sixty-five-percent-revenue-growth", "date": "2026-01-27" } }, { "text": "Daniela Braga founded the company in 2015. It has raised more than $85 million and holds ISO 42001 certification.", "source": { "label": "Defined.ai, awarded ISO 42001 certification", "href": "https://defined.ai/press-room/defined-ai-awarded-iso-42001-certification", "date": "2026-04-28" } } ], "terms": { "whatTheyBuy": { "label": "What they license", "value": "Audio, video, image, text or multimodal data; cinematic, anime, coding and STEM data in high demand", "source": { "label": "Defined.ai, partnership programs", "href": "https://defined.ai/partnership-programs", "date": "2026-10-17" } }, "priceOrBands": { "label": "Price or bands", "value": "Not published", "source": null }, "headcountFloor": { "label": "Headcount floor", "value": "None stated; any business that generates or aggregates data qualifies", "source": { "label": "Defined.ai, partnership programs", "href": "https://defined.ai/partnership-programs", "date": "2026-10-17" } }, "historyFloor": { "label": "Years of history", "value": "Not published", "source": null }, "geography": { "label": "Geography", "value": "Not published", "source": null }, "exclusivity": { "label": "Exclusivity", "value": "Not published", "source": null }, "paymentTiming": { "label": "Payment timing", "value": "Not published", "source": null }, "whoDeidentifies": { "label": "Who de-identifies", "value": "Not published", "source": null }, "referralBounty": { "label": "Referral bounty", "value": "Not published", "source": null }, "publishedTermsPage": { "label": "Published terms page", "value": "Not published", "source": null } }, "sources": [ { "label": "Defined.ai, partnership programs", "href": "https://defined.ai/partnership-programs", "date": "2026-10-17" }, { "label": "Defined.ai, data license agreement (buyer side, effective Nov 30, 2022)", "href": "https://defined.ai/data-license-agreement", "date": "2026-10-17" }, { "label": "Defined.ai, reports 65% revenue growth", "href": "https://defined.ai/press-room/defined-ai-reports-sixty-five-percent-revenue-growth", "date": "2026-01-27" }, { "label": "Defined.ai, awarded ISO 42001 certification", "href": "https://defined.ai/press-room/defined-ai-awarded-iso-42001-certification", "date": "2026-04-28" }, { "label": "The Decoder, on Reuters report of AI training data prices", "href": "https://the-decoder.com/tech-giants-are-on-a-billion-dollar-shopping-spree-for-ai-training-data/", "date": "2024-04-06" }, { "label": "Who Sells Data, buying programs register", "href": "https://whosellsdata.com/buying", "date": "2026-10-14" } ] }, { "slug": "endless", "name": "Endless", "kind": "Data partner", "url": "https://marindata.ai/buyers/endless", "summary": "Endless licenses the working records of operating companies to AI labs for training and evaluation. It agrees a fixed sum before anything is shared and keeps licenses non-exclusive.", "goodFitFor": "Companies of about 30 or more staff in the US, UK or Canada with years of quotes, tickets and work orders who want to keep the right to license elsewhere.", "watchFor": [ "The up to $4M figure is a headline; Endless shows a company its own range only after it enters a work email (Endless, observed Oct 19, 2026).", "Endless is a trade name of Seldon Ventures, L.L.C., a Delaware company in Newark, Delaware (Endless, observed Oct 19, 2026)." ], "lastVerified": "2026-11-02", "termsPublished": 9, "notes": [], "terms": { "whatTheyBuy": { "label": "What they license", "value": "Quotes, tickets, work orders, SOPs, handover notes and the decisions behind them", "source": { "label": "Endless, homepage", "href": "https://endless-data.com/", "date": "2026-10-19" } }, "priceOrBands": { "label": "Price or bands", "value": "Headline of up to $4M; the payout range appears only after entering a work email", "source": { "label": "Endless, homepage", "href": "https://endless-data.com/", "date": "2026-10-19" } }, "headcountFloor": { "label": "Headcount floor", "value": "Roughly 30 or more employees", "source": { "label": "Endless, homepage", "href": "https://endless-data.com/", "date": "2026-10-19" } }, "historyFloor": { "label": "Years of history", "value": "Several years of history", "source": { "label": "Endless, homepage", "href": "https://endless-data.com/", "date": "2026-10-19" } }, "geography": { "label": "Geography", "value": "US, UK and Canada prioritised", "source": { "label": "Endless, homepage", "href": "https://endless-data.com/", "date": "2026-10-19" } }, "exclusivity": { "label": "Exclusivity", "value": "Non-exclusive; the company can license the same data to others", "source": { "label": "Endless, homepage", "href": "https://endless-data.com/", "date": "2026-10-19" } }, "paymentTiming": { "label": "Payment timing", "value": "Fixed sum agreed before anything is shared; paid after transfer", "source": { "label": "Endless, homepage", "href": "https://endless-data.com/", "date": "2026-10-19" } }, "whoDeidentifies": { "label": "Who de-identifies", "value": "Endless, before anything leaves the company", "source": { "label": "Endless, homepage", "href": "https://endless-data.com/", "date": "2026-10-19" } }, "referralBounty": { "label": "Referral bounty", "value": "Not published", "source": null }, "publishedTermsPage": { "label": "Published terms page", "value": "Terms and Conditions at /terms", "source": { "label": "Endless, homepage", "href": "https://endless-data.com/", "date": "2026-10-19" } } }, "sources": [ { "label": "Endless, homepage", "href": "https://endless-data.com/", "date": "2026-10-19" } ] }, { "slug": "google-spirit-airlines", "name": "Google (Spirit Airlines data)", "kind": "AI lab", "url": "https://marindata.ai/buyers/google-spirit-airlines", "summary": "Google won the August 2026 bankruptcy auction for Spirit Airlines' internal business data with a $10M bid. Separately, Google runs an invite-only pilot that pays for the delivery of non-public content.", "goodFitFor": "Very large record sets sold through a court process or an invited partnership. Google has no public application route.", "watchFor": [ "Court approval was still pending. The hearing moved from Oct 14 to Nov 6, 2026 after more than 120 lawmakers raised concerns (TravelPulse, Oct 10, 2026).", "Unions for Spirit's flight attendants and pilots objected, as did the software maker Springshot. The privacy ombudsman recommended approval on Oct 5, 2026 after passenger databases were taken out of the deal (The Next Web, Oct 9, 2026).", "micro1 filed a late $12.5M bid after the auction, and Mercor was the backup bidder at $7.5M (ch-aviation, Aug 24, 2026)." ], "lastVerified": "2026-11-02", "termsPublished": 3, "notes": [], "terms": { "whatTheyBuy": { "label": "What they license", "value": "Spirit's employee emails, Microsoft Teams messages, spreadsheets, calendars and operations data, for product development and AI training", "source": { "label": "iTnews (Reuters), Google to buy Spirit Airlines business data", "href": "https://www.itnews.com.au/news/google-to-buy-spirit-airlines-business-data-for-us10-million-628253", "date": "2026-08-19" } }, "priceOrBands": { "label": "Price or bands", "value": "$10M for about 100 million emails, 500 million Teams messages and 176,000 employee records", "source": { "label": "MIT Technology Review, AI labs are buying data from dead startups", "href": "https://technologyreview.com/2026/10/13/1145755/ai-labs-buying-data-from-dead-startups-to-train-models", "date": "2026-10-13" } }, "headcountFloor": { "label": "Headcount floor", "value": "Not published", "source": null }, "historyFloor": { "label": "Years of history", "value": "Not published", "source": null }, "geography": { "label": "Geography", "value": "Not published", "source": null }, "exclusivity": { "label": "Exclusivity", "value": "Not published", "source": null }, "paymentTiming": { "label": "Payment timing", "value": "Not published", "source": null }, "whoDeidentifies": { "label": "Who de-identifies", "value": "Not published", "source": null }, "referralBounty": { "label": "Referral bounty", "value": "Not published", "source": null }, "publishedTermsPage": { "label": "Published terms page", "value": "Invite-only content partnership page; no public application route", "source": { "label": "Google, partnerships to improve our AI products", "href": "https://ai.google/partnerships-to-improve-our-ai-products/", "date": "2026-10-14" } } }, "sources": [ { "label": "iTnews (Reuters), Google to buy Spirit Airlines business data", "href": "https://www.itnews.com.au/news/google-to-buy-spirit-airlines-business-data-for-us10-million-628253", "date": "2026-08-19" }, { "label": "MIT Technology Review, AI labs are buying data from dead startups", "href": "https://technologyreview.com/2026/10/13/1145755/ai-labs-buying-data-from-dead-startups-to-train-models", "date": "2026-10-13" }, { "label": "TravelPulse, Spirit data sale delayed as lawmakers raise concerns", "href": "https://www.travelpulse.com/news/airlines-airports/spirit-airlines-google-10-million-data-sale-delayed-as-lawmakers-raise-concerns", "date": "2026-10-10" }, { "label": "The Next Web, Spirit's privacy ombudsman backs Google's data deal", "href": "https://thenextweb.com/news/google-spirit-airlines-data-ombudsman-approval", "date": "2026-10-09" }, { "label": "ch-aviation, court delays Google acquisition of Spirit data", "href": "https://www.ch-aviation.com/news/170908-court-delays-googles-acquisition-of-spirits-business-data", "date": "2026-08-24" }, { "label": "Google, partnerships to improve our AI products", "href": "https://ai.google/partnerships-to-improve-our-ai-products/", "date": "2026-10-14" } ] }, { "slug": "grepped", "name": "Grepped", "kind": "Intermediary", "url": "https://marindata.ai/buyers/grepped", "summary": "Grepped AI, Inc. is a San Francisco company founded in 2026 that packages company data and lists it to AI labs. It connects to work tools read-only and says the money goes to the owner when a lab buys.", "goodFitFor": "Owners who want a fast, self-serve listing and are able to judge offers themselves.", "watchFor": [ "Grepped does not publish what it keeps from a deal (grepped.ai, observed Oct 16, 2026).", "Its reported $4M MRR, a $48M annual run rate, two weeks after launch was announced by the founder on X and is unverified (ARR Club, Oct 6, 2026).", "Referrers agree not to help a referred owner license its data outside Grepped for 12 months after the terms end (Grepped referral terms, clause 5.5)." ], "lastVerified": "2026-11-02", "termsPublished": 6, "notes": [], "terms": { "whatTheyBuy": { "label": "What they license", "value": "Messages, codebases, documents, CRM and finance records, and task records from tools such as Slack, GitHub, Drive, Salesforce, QuickBooks and Zendesk", "source": { "label": "Grepped, homepage", "href": "https://grepped.ai", "date": "2026-10-16" } }, "priceOrBands": { "label": "Price or bands", "value": "Homepage offers \"premium licensing for company data worth millions\" and \"up to 6 figures per workspace\"; the earlier \"$20K–$5M\" estimate is no longer on the page", "source": { "label": "Grepped, homepage", "href": "https://www.grepped.ai/", "date": "2026-10-26" } }, "headcountFloor": { "label": "Headcount floor", "value": "Not published", "source": null }, "historyFloor": { "label": "Years of history", "value": "Not published", "source": null }, "geography": { "label": "Geography", "value": "Not published", "source": null }, "exclusivity": { "label": "Exclusivity", "value": "Not published", "source": null }, "paymentTiming": { "label": "Payment timing", "value": "\"As little as 7 days,\" per the homepage", "source": { "label": "Grepped, homepage", "href": "https://grepped.ai", "date": "2026-10-16" } }, "whoDeidentifies": { "label": "Who de-identifies", "value": "Grepped strips customer names, contact details and identifying information; buyers see redacted previews", "source": { "label": "Grepped, homepage", "href": "https://grepped.ai", "date": "2026-10-16" } }, "referralBounty": { "label": "Referral bounty", "value": "10% of the owner payment, capped at $100K per deal, paid when the owner is paid and within 30 days of the lab paying", "source": { "label": "Grepped referral terms (effective Oct 13, 2026)", "href": "https://www.grepped.ai/assets/grepped-referral-terms.pdf", "date": "2026-10-13" } }, "publishedTermsPage": { "label": "Published terms page", "value": "Referral terms published (effective Oct 13, 2026); seller fee not published", "source": { "label": "Grepped referral terms (effective Oct 13, 2026)", "href": "https://www.grepped.ai/assets/grepped-referral-terms.pdf", "date": "2026-10-13" } } }, "sources": [ { "label": "Grepped, homepage", "href": "https://grepped.ai", "date": "2026-10-16" }, { "label": "Grepped, homepage", "href": "https://www.grepped.ai/", "date": "2026-10-26" }, { "label": "Grepped referral terms (effective Oct 13, 2026)", "href": "https://www.grepped.ai/assets/grepped-referral-terms.pdf", "date": "2026-10-13" }, { "label": "ARR Club, Grepped run-rate claim", "href": "https://www.arr.club/signal/grepped-hit-48m-arr-run-rate-two-weeks-after-launch", "date": "2026-10-06" } ] }, { "slug": "handshake-ai", "name": "Handshake AI", "kind": "Data partner", "url": "https://marindata.ai/buyers/handshake-ai", "summary": "Handshake AI licenses scoped, de-identified company data such as business process flows, hiring logic and document-level communications. It publishes a range of $100K to $4M+.", "goodFitFor": "Companies of 20 or more full-time staff with three or more years in operation and data across common business tools.", "watchFor": [ "The license page said $100K to $4M on Oct 14, 2026 and says $100K to $4M+ on Oct 19, 2026 (Handshake, Who Sells Data).", "No separate referral terms are published, so rules on who can refer and on clawbacks are not stated (Who Sells Data, Oct 14, 2026)." ], "lastVerified": "2026-11-02", "termsPublished": 6, "notes": [], "terms": { "whatTheyBuy": { "label": "What they license", "value": "Operational data the company chooses to share, from sources such as Gmail, Slack, Google Drive, NetSuite, QuickBooks and AutoCAD", "source": { "label": "Handshake AI, data partnerships license page", "href": "https://joinhandshake.com/ai/data-partnerships/license/", "date": "2026-10-16" } }, "priceOrBands": { "label": "Price or bands", "value": "$100K to $4M+, depending on sources, volume, quality, due diligence and license terms", "source": { "label": "Handshake AI, data partnerships license page", "href": "https://joinhandshake.com/ai/data-partnerships/license/", "date": "2026-10-19" } }, "headcountFloor": { "label": "Headcount floor", "value": "20+ full-time employees", "source": { "label": "Handshake AI, data partnerships license page", "href": "https://joinhandshake.com/ai/data-partnerships/license/", "date": "2026-10-16" } }, "historyFloor": { "label": "Years of history", "value": "3+ years in operation", "source": { "label": "Handshake AI, data partnerships license page", "href": "https://joinhandshake.com/ai/data-partnerships/license/", "date": "2026-10-16" } }, "geography": { "label": "Geography", "value": "Not published", "source": null }, "exclusivity": { "label": "Exclusivity", "value": "Not published", "source": null }, "paymentTiming": { "label": "Payment timing", "value": "Typically 30–60 days after Handshake accepts the de-identified dataset", "source": { "label": "Who Sells Data, buying programs register", "href": "https://whosellsdata.com/buying", "date": "2026-10-14" } }, "whoDeidentifies": { "label": "Who de-identifies", "value": "Not published", "source": null }, "referralBounty": { "label": "Referral bounty", "value": "$10K per referred company that signs, paid when the deal closes", "source": { "label": "Handshake AI, referral", "href": "https://joinhandshake.com/ai/data-partnerships/referral/", "date": "2026-10-14" } }, "publishedTermsPage": { "label": "Published terms page", "value": "Not published", "source": null } }, "sources": [ { "label": "Handshake AI, data partnerships license page", "href": "https://joinhandshake.com/ai/data-partnerships/license/", "date": "2026-10-16" }, { "label": "Handshake AI, data partnerships license page", "href": "https://joinhandshake.com/ai/data-partnerships/license/", "date": "2026-10-19" }, { "label": "Handshake AI, referral", "href": "https://joinhandshake.com/ai/data-partnerships/referral/", "date": "2026-10-14" }, { "label": "Who Sells Data, buying programs register", "href": "https://whosellsdata.com/buying", "date": "2026-10-14" } ] }, { "slug": "hud-datavendor", "name": "HUD (DataVendor)", "kind": "Marketplace", "url": "https://marindata.ai/buyers/hud-datavendor", "summary": "HUD runs DataVendor, a beta marketplace where organizations sell codebases, records, tasksets and RL environments to invited buyers. HUD keeps a 20% fee when a vendor withdraws to a bank.", "goodFitFor": "Organizations with codebases or tasksets that want to list them themselves and skip a negotiated deal.", "watchFor": [ "DataVendor is in beta, and its fees are current settings that can change (Who Sells Data, Oct 14, 2026).", "HUD's own comparison of codebase platforms ranks HUD first (HUD, Jun 24, 2026)." ], "lastVerified": "2026-11-02", "termsPublished": 5, "notes": [], "terms": { "whatTheyBuy": { "label": "What they license", "value": "Codebases, raw data such as documents, logs and enterprise records, tasksets, and RL environments sold with a HUD taskset", "source": { "label": "HUD DataVendor, how it works", "href": "https://docs.datavendor.ai/overview/how-it-works", "date": "2026-10-14" } }, "priceOrBands": { "label": "Price or bands", "value": "Sale proceeds less a 20% fee on bank withdrawals; HUD cites a $5,000 baseline per codebase", "source": { "label": "HUD, best platforms for selling a codebase", "href": "https://www.hud.ai/resources/best-platforms-selling-codebase", "date": "2026-06-24" } }, "headcountFloor": { "label": "Headcount floor", "value": "Organizations only; no headcount floor", "source": { "label": "HUD DataVendor, how it works", "href": "https://docs.datavendor.ai/overview/how-it-works", "date": "2026-10-14" } }, "historyFloor": { "label": "Years of history", "value": "Not published", "source": null }, "geography": { "label": "Geography", "value": "Not published", "source": null }, "exclusivity": { "label": "Exclusivity", "value": "Not published", "source": null }, "paymentTiming": { "label": "Payment timing", "value": "Earnings settle 14 days after payment; withdrawals need $1,000 before fees", "source": { "label": "HUD DataVendor, how it works", "href": "https://docs.datavendor.ai/overview/how-it-works", "date": "2026-10-14" } }, "whoDeidentifies": { "label": "Who de-identifies", "value": "Not published", "source": null }, "referralBounty": { "label": "Referral bounty", "value": "Not published", "source": null }, "publishedTermsPage": { "label": "Published terms page", "value": "How-it-works documentation with current fee settings", "source": { "label": "HUD DataVendor, how it works", "href": "https://docs.datavendor.ai/overview/how-it-works", "date": "2026-10-14" } } }, "sources": [ { "label": "HUD DataVendor, how it works", "href": "https://docs.datavendor.ai/overview/how-it-works", "date": "2026-10-14" }, { "label": "HUD, best platforms for selling a codebase", "href": "https://www.hud.ai/resources/best-platforms-selling-codebase", "date": "2026-06-24" }, { "label": "Who Sells Data, buying programs register", "href": "https://whosellsdata.com/buying", "date": "2026-10-14" } ] }, { "slug": "labelbox-prism", "name": "Labelbox Prism", "kind": "Data partner", "url": "https://marindata.ai/buyers/labelbox-prism", "summary": "Prism is the data licensing program of Labelbox, an AI data company. It licenses real-world operational data from companies and individual contributors, and its licenses are exclusive by default.", "goodFitFor": "Companies willing to grant exclusivity in exchange for an upfront fee and a share of later use.", "watchFor": [ "Licenses are exclusive by default (Prism, observed Oct 16, 2026). Settle scope before signing if you plan to license the same records elsewhere." ], "lastVerified": "2026-11-02", "termsPublished": 5, "notes": [], "terms": { "whatTheyBuy": { "label": "What they license", "value": "Work documents and operational data from systems such as Salesforce, Jira, SAP and Slack", "source": { "label": "Labelbox Prism", "href": "https://prism.labelbox.com/", "date": "2026-10-16" } }, "priceOrBands": { "label": "Price or bands", "value": "About $50K for smaller targeted datasets to $1M+ for large enterprise partnerships", "source": { "label": "Labelbox Prism", "href": "https://prism.labelbox.com/", "date": "2026-10-16" } }, "headcountFloor": { "label": "Headcount floor", "value": "Not published", "source": null }, "historyFloor": { "label": "Years of history", "value": "Not published", "source": null }, "geography": { "label": "Geography", "value": "Not published", "source": null }, "exclusivity": { "label": "Exclusivity", "value": "Exclusive by default, tailored to the scope, sensitivity and value of the data", "source": { "label": "Labelbox Prism", "href": "https://prism.labelbox.com/", "date": "2026-10-16" } }, "paymentTiming": { "label": "Payment timing", "value": "Upfront fee at signing plus a revenue share on later use; Prism cites under 21 days from first call to first wire", "source": { "label": "Labelbox Prism", "href": "https://prism.labelbox.com/", "date": "2026-10-16" } }, "whoDeidentifies": { "label": "Who de-identifies", "value": "Prism handles ingestion, normalization and PII scrubbing", "source": { "label": "Labelbox Prism", "href": "https://prism.labelbox.com/", "date": "2026-10-16" } }, "referralBounty": { "label": "Referral bounty", "value": "Not published", "source": null }, "publishedTermsPage": { "label": "Published terms page", "value": "Not published", "source": null } }, "sources": [ { "label": "Labelbox Prism", "href": "https://prism.labelbox.com/", "date": "2026-10-16" }, { "label": "Who Sells Data, buying programs register", "href": "https://whosellsdata.com/buying", "date": "2026-10-14" } ] }, { "slug": "mercor", "name": "Mercor", "kind": "Data partner", "url": "https://marindata.ai/buyers/mercor", "summary": "Mercor is an AI data and expert-talent company. Mercor Data licenses workflow data pulled from more than 100 business tools and pays the company when the anonymized dataset reaches a buyer.", "goodFitFor": "Companies in legal, healthcare, finance and insurance, professional services, manufacturing, retail, real estate or logistics whose work runs through common cloud tools.", "watchFor": [ "Mercor publishes no price. It says pay depends on data volume, the tools connected and data depth (Mercor, observed Oct 16, 2026).", "In April 2026 Mercor confirmed a security breach tied to a supply-chain attack on the LiteLLM library. The extortion group Lapsus$ claimed up to 4TB of data, which Mercor had not confirmed (Fortune, Apr 2, 2026).", "Anonymization is fully automated, with no manual review (Mercor FAQ, observed Oct 16, 2026)." ], "lastVerified": "2026-11-02", "termsPublished": 5, "notes": [ { "text": "In April 2026 Mercor confirmed a breach it tied to a supply-chain attack on LiteLLM, an open-source library. It said it was one of thousands of companies affected and that the malicious code was removed within hours of discovery.", "source": { "label": "Fortune, Mercor security incident", "href": "https://fortune.com/2026/04/02/mercor-ai-startup-security-incident-10-billion/", "date": "2026-04-02" } }, { "text": "The extortion group Lapsus$ claimed up to four terabytes of data, including source code, database records, Slack data and internal tickets. Unconfirmed reports said datasets from some Mercor customers and details of their confidential AI projects may have been taken. Mercor had not confirmed those reports.", "source": { "label": "Fortune, Mercor security incident", "href": "https://fortune.com/2026/04/02/mercor-ai-startup-security-incident-10-billion/", "date": "2026-04-02" } }, { "text": "The class actions that followed were brought by Mercor's contractors and job candidates, not by client companies. The Wall Street Journal counted at least seven by late April 2026. Mercor said it strongly disputes the claims.", "source": { "label": "Jerusalem Post, AI firm Mercor faces lawsuits over alleged data breach", "href": "https://www.jpost.com/business-and-innovation/article-894282", "date": "2026-04-27" } }, { "text": "One of them, Ananthula v. Mercor.io Corporation, No. 3:26-cv-03362 (N.D. Cal., filed Apr 21, 2026), alleges the theft of recorded interviews, facial biometric data, background checks and tax and banking details, plus about 939 GB of source code and internal systems data. These are the plaintiffs' allegations, not findings.", "source": { "label": "Hausfeld (plaintiffs' counsel), Mercor data breach case page", "href": "https://www.hausfeld.com/what-we-do/current-claims/mercor-data-breach", "date": "2026-10-19" } } ], "terms": { "whatTheyBuy": { "label": "What they license", "value": "Data from business tools such as Slack, Teams, Gmail, Outlook, Confluence, Salesforce, GitHub, Drive and Notion", "source": { "label": "Mercor Data", "href": "https://www.mercor.com/data/", "date": "2026-10-16" } }, "priceOrBands": { "label": "Price or bands", "value": "Not published", "source": null }, "headcountFloor": { "label": "Headcount floor", "value": "Not published", "source": null }, "historyFloor": { "label": "Years of history", "value": "Not published", "source": null }, "geography": { "label": "Geography", "value": "Not published", "source": null }, "exclusivity": { "label": "Exclusivity", "value": "Not published", "source": null }, "paymentTiming": { "label": "Payment timing", "value": "By wire once the anonymized dataset is delivered to the buyer, typically 2–4 weeks after signing", "source": { "label": "Mercor Data", "href": "https://www.mercor.com/data/", "date": "2026-10-16" } }, "whoDeidentifies": { "label": "Who de-identifies", "value": "Mercor, through a fully automated pipeline; its engineers do not review the data by hand", "source": { "label": "Mercor Data", "href": "https://www.mercor.com/data/", "date": "2026-10-16" } }, "referralBounty": { "label": "Referral bounty", "value": "A bonus is paid when the referred company goes live and starts earning; the amount is not on the public page", "source": { "label": "Mercor Data", "href": "https://www.mercor.com/data/", "date": "2026-10-16" } }, "publishedTermsPage": { "label": "Published terms page", "value": "Program page with process and FAQ at mercor.com/data", "source": { "label": "Mercor Data", "href": "https://www.mercor.com/data/", "date": "2026-10-16" } } }, "sources": [ { "label": "Mercor Data", "href": "https://www.mercor.com/data/", "date": "2026-10-16" }, { "label": "Fortune, Mercor security incident", "href": "https://fortune.com/2026/04/02/mercor-ai-startup-security-incident-10-billion/", "date": "2026-04-02" }, { "label": "Jerusalem Post, AI firm Mercor faces lawsuits over alleged data breach", "href": "https://www.jpost.com/business-and-innovation/article-894282", "date": "2026-04-27" }, { "label": "Hausfeld (plaintiffs' counsel), Mercor data breach case page", "href": "https://www.hausfeld.com/what-we-do/current-claims/mercor-data-breach", "date": "2026-10-19" }, { "label": "aitrainer.work, AI labs paying for enterprise workflow data", "href": "https://aitrainer.work/news/ai-labs-paying-enterprise-workflow-data-2026", "date": "2026-06-29" }, { "label": "ch-aviation, court delays Google acquisition of Spirit data", "href": "https://www.ch-aviation.com/news/170908-court-delays-googles-acquisition-of-spirits-business-data", "date": "2026-08-24" }, { "label": "Who Sells Data, buying programs register", "href": "https://whosellsdata.com/buying", "date": "2026-10-14" } ] }, { "slug": "micro1", "name": "micro1", "kind": "Data partner", "url": "https://marindata.ai/buyers/micro1", "summary": "micro1 builds AI training environments from de-identified company data and sells them to frontier labs. Its Enterprise Data Partnership licenses operational data from established companies, and on Oct 9, 2026 micro1 committed $1B over 12 months to the program.", "goodFitFor": "US companies with 30 or more staff, mature documentation and English-language records spread across several teams.", "watchFor": [ "aitrainer.work (Jun 29, 2026) reported the tiers as yearly amounts. micro1's own partnership page gives them with no period attached.", "micro1's company-referral page lists $100K–$2M+ for approved data packages, a wider range than the partnership tiers (micro1, Oct 26, 2026).", "A referral bonus is not earned when a company registers or signs. The data must be transferred and approved first, and only referrals made through the registered link are credited (micro1, Oct 26, 2026)." ], "lastVerified": "2026-11-02", "termsPublished": 8, "notes": [ { "text": "micro1 says it licenses data only to labs and enterprises based in the U.S. and allied countries, and keeps each company's data segregated from other companies' data. It also buys AI agent trajectories, which it says qualify for $1–2M annually.", "source": { "label": "micro1, data partnerships", "href": "https://www.micro1.ai/data-partnerships", "date": "2026-10-26" } }, { "text": "A data broker, Ryan Lock, said on the NGMI podcast in October 2026 that micro1 pays $50K–$75K referral fees. micro1's referral page now publishes $50K for enterprise partnerships and $75K for defunct companies, so his figure matches the published terms.", "source": { "label": "NGMI podcast, interview with Ryan Lock", "href": "https://podcasts.apple.com/us/podcast/ngmi/id1895066904", "date": "2026-10-16" } } ], "terms": { "whatTheyBuy": { "label": "What they license", "value": "SOPs and process documentation, internal knowledge bases, CRM records, support tickets, internal communications, project histories, workflows and templates; also AI agent trajectories", "source": { "label": "micro1, data partnerships", "href": "https://www.micro1.ai/data-partnerships", "date": "2026-10-26" } }, "priceOrBands": { "label": "Price or bands", "value": "Published tiers of $100K-$500K, $500K-$2M and $2M+, set by the quality, uniqueness and value of the data; the headline reads \"$200k-2M+\" and payments \"typically ranging from $150K to over $2M\"", "source": { "label": "micro1, data partnerships", "href": "https://www.micro1.ai/data-partnerships", "date": "2026-10-26" } }, "headcountFloor": { "label": "Headcount floor", "value": "30+ employees", "source": { "label": "micro1, data partnerships", "href": "https://www.micro1.ai/data-partnerships", "date": "2026-10-26" } }, "historyFloor": { "label": "Years of history", "value": "Not published", "source": null }, "geography": { "label": "Geography", "value": "US first, then other Western markets; demand is strongest in the US, UK and Canada; English documentation required", "source": { "label": "micro1, data partnerships", "href": "https://www.micro1.ai/data-partnerships", "date": "2026-10-26" } }, "exclusivity": { "label": "Exclusivity", "value": "Not published", "source": null }, "paymentTiming": { "label": "Payment timing", "value": "\"In full within 30 days of acceptance\"; the FAQ says micro1 has 30 days to review and approve the data, then payment follows within 30 days of invoicing", "source": { "label": "micro1, data partnerships", "href": "https://www.micro1.ai/data-partnerships", "date": "2026-10-26" } }, "whoDeidentifies": { "label": "Who de-identifies", "value": "micro1 removes or transforms personal and identifying information before the data goes to third parties; the company does not have to anonymize it first", "source": { "label": "micro1, data partnerships", "href": "https://www.micro1.ai/data-partnerships", "date": "2026-10-26" } }, "referralBounty": { "label": "Referral bounty", "value": "Enterprise data partnerships: $50K for deals up to $700K, 7.5% of the deal value above that. Defunct companies: $75K for deals up to $1M, 7.5% above. Other deals: 7.5%. Paid by wire after micro1 approves the transferred data", "source": { "label": "micro1, company referral", "href": "https://www.micro1.ai/company-referral", "date": "2026-10-26" } }, "publishedTermsPage": { "label": "Published terms page", "value": "Yes: a data partnership page and company referral terms", "source": { "label": "micro1, company referral", "href": "https://www.micro1.ai/company-referral", "date": "2026-10-26" } } }, "sources": [ { "label": "micro1, data partnerships", "href": "https://www.micro1.ai/data-partnerships", "date": "2026-10-26" }, { "label": "micro1, company referral", "href": "https://www.micro1.ai/company-referral", "date": "2026-10-26" }, { "label": "micro1, $1B enterprise data commitment", "href": "https://www.micro1.ai/blog/micro1-commits-1b-to-enterprise-data-acquisition", "date": "2026-10-09" }, { "label": "aitrainer.work, AI labs paying for enterprise workflow data", "href": "https://aitrainer.work/news/ai-labs-paying-enterprise-workflow-data-2026", "date": "2026-06-29" }, { "label": "Who Sells Data, buying programs register", "href": "https://whosellsdata.com/buying", "date": "2026-10-14" } ] }, { "slug": "mode", "name": "Mode (data.mode.inc)", "kind": "Data partner", "url": "https://marindata.ai/buyers/mode", "summary": "Mode, at data.mode.inc, buys an agreed copy of company operating records directly and charges no broker fee. Its main business pays consumers for personal data; this entry covers only the company program.", "goodFitFor": "US firms with years of owned records, including accounting firms of 10 or more and law firms of 6 or more that other programs turn away.", "watchFor": [ "Mode's page listed 30+ employees for most businesses on Oct 16, 2026. The Who Sells Data register recorded 20+ US office employees on Oct 14, 2026.", "Mode labels the calculator range indicative and not a guaranteed offer (Mode FAQ, observed Oct 16, 2026).", "A referral counts only if the company signs a full corpus license within 6 months (Who Sells Data, Oct 8, 2026)." ], "lastVerified": "2026-11-02", "termsPublished": 9, "notes": [ { "text": "Mode also markets the program to law firms as BizBuyData. Above the Law reported the launch on Oct 16, 2026, and bizbuydata.com redirects to data.mode.inc.", "source": { "label": "Above the Law, tech company launches program to pay law firms for their emails", "href": "https://abovethelaw.com/2026/10/tech-company-launches-program-to-pay-law-firms-for-their-emails-privilege-be-damned/", "date": "2026-10-16" } }, { "text": "Mode told Above the Law that its tools remove sensitive information and replace names and other identifying details before records are used.", "source": { "label": "Above the Law, tech company launches program to pay law firms for their emails", "href": "https://abovethelaw.com/2026/10/tech-company-launches-program-to-pay-law-firms-for-their-emails-privilege-be-damned/", "date": "2026-10-16" } } ], "terms": { "whatTheyBuy": { "label": "What they license", "value": "Email, documents, chat, calls and transcripts, CRM and billing or operations data, with systems and date ranges agreed before export", "source": { "label": "Mode, company data", "href": "https://data.mode.inc/", "date": "2026-10-16" } }, "priceOrBands": { "label": "Price or bands", "value": "Indicative $100K–$5M; a fixed offer follows a review of the records", "source": { "label": "Mode, company data", "href": "https://data.mode.inc/", "date": "2026-10-16" } }, "headcountFloor": { "label": "Headcount floor", "value": "30+ employees for most businesses; 10+ for accounting, finance and medical firms; 6+ for law firms", "source": { "label": "Mode, company data", "href": "https://data.mode.inc/", "date": "2026-10-16" } }, "historyFloor": { "label": "Years of history", "value": "Several years of company-owned records (strongest fit)", "source": { "label": "Mode, company data", "href": "https://data.mode.inc/", "date": "2026-10-16" } }, "geography": { "label": "Geography", "value": "Teams primarily based in the US (strongest fit)", "source": { "label": "Mode, company data", "href": "https://data.mode.inc/", "date": "2026-10-16" } }, "exclusivity": { "label": "Exclusivity", "value": "Not published", "source": null }, "paymentTiming": { "label": "Payment timing", "value": "About three months from first conversation to payment", "source": { "label": "Mode, company data", "href": "https://data.mode.inc/", "date": "2026-10-16" } }, "whoDeidentifies": { "label": "Who de-identifies", "value": "Mode, in a secure workspace, before onward delivery", "source": { "label": "Mode, company data", "href": "https://data.mode.inc/", "date": "2026-10-16" } }, "referralBounty": { "label": "Referral bounty", "value": "6% of the deal or $50K, whichever is greater; 6% with no minimum for companies based outside the US; paid within 30 days of Mode's payment; 12-month clawback", "source": { "label": "Who Sells Data, Mode referral entry", "href": "https://whosellsdata.com/buying", "date": "2026-10-08" } }, "publishedTermsPage": { "label": "Published terms page", "value": "Affiliate Referral Agreement v1.0, published Oct 6, 2026", "source": { "label": "Who Sells Data, Mode referral entry", "href": "https://whosellsdata.com/buying", "date": "2026-10-08" } } }, "sources": [ { "label": "Mode, company data", "href": "https://data.mode.inc/", "date": "2026-10-16" }, { "label": "Who Sells Data, Mode referral entry", "href": "https://whosellsdata.com/buying", "date": "2026-10-08" }, { "label": "Who Sells Data, buying programs register", "href": "https://whosellsdata.com/buying", "date": "2026-10-14" }, { "label": "Above the Law, tech company launches program to pay law firms for their emails", "href": "https://abovethelaw.com/2026/10/tech-company-launches-program-to-pay-law-firms-for-their-emails-privilege-be-damned/", "date": "2026-10-16" }, { "label": "BizBuyData.com (redirects to data.mode.inc)", "href": "https://bizbuydata.com/", "date": "2026-10-19" } ] }, { "slug": "nyne-ai", "name": "Nyne AI", "kind": "Data partner", "url": "https://marindata.ai/buyers/nyne-ai", "summary": "Nyne AI licenses company workflow data from chat, email, documents and business platforms. It also sells people-data APIs; this entry covers only its company data program.", "goodFitFor": "US companies with 20 or more W-2 staff and three or more years of workflows in common tools.", "watchFor": [ "Data is transferred as it exists and de-identified on Nyne's side, so raw records leave the company (Nyne, observed Oct 16, 2026).", "Nyne's estimator showed $40,658–$95,198 at its default inputs on Oct 19, 2026 ($44,379–$103,910 on Oct 16). It labels the estimate indicative and not an offer." ], "lastVerified": "2026-11-02", "termsPublished": 8, "notes": [], "terms": { "whatTheyBuy": { "label": "What they license", "value": "Chat, email and calendar, documents and files, and business platforms such as Salesforce, HubSpot, Workday, ServiceNow, SAP and Zendesk", "source": { "label": "Nyne AI, data partnerships", "href": "https://nyne.ai/data-partnerships", "date": "2026-10-16" } }, "priceOrBands": { "label": "Price or bands", "value": "\"Six-figure payouts\"", "source": { "label": "Who Sells Data, buying programs register", "href": "https://whosellsdata.com/buying", "date": "2026-10-14" } }, "headcountFloor": { "label": "Headcount floor", "value": "20+ full-time US W-2 employees", "source": { "label": "Nyne AI, data partnerships", "href": "https://nyne.ai/data-partnerships", "date": "2026-10-16" } }, "historyFloor": { "label": "Years of history", "value": "3+ years of operational workflows", "source": { "label": "Nyne AI, data partnerships", "href": "https://nyne.ai/data-partnerships", "date": "2026-10-16" } }, "geography": { "label": "Geography", "value": "US (W-2 employees)", "source": { "label": "Nyne AI, data partnerships", "href": "https://nyne.ai/data-partnerships", "date": "2026-10-16" } }, "exclusivity": { "label": "Exclusivity", "value": "Not published", "source": null }, "paymentTiming": { "label": "Payment timing", "value": "After the dataset is delivered and accepted, as set out in the partnership agreement", "source": { "label": "Nyne AI, data partnerships", "href": "https://nyne.ai/data-partnerships", "date": "2026-10-16" } }, "whoDeidentifies": { "label": "Who de-identifies", "value": "Nyne, on its side, after the data is transferred as it exists", "source": { "label": "Nyne AI, data partnerships", "href": "https://nyne.ai/data-partnerships", "date": "2026-10-16" } }, "referralBounty": { "label": "Referral bounty", "value": "Up to $20K per introduced company that completes a transaction; no written terms, referrals by email", "source": { "label": "Nyne AI, referrals", "href": "https://nyne.ai/referrals", "date": "2026-10-14" } }, "publishedTermsPage": { "label": "Published terms page", "value": "Not published", "source": null } }, "sources": [ { "label": "Nyne AI, data partnerships", "href": "https://nyne.ai/data-partnerships", "date": "2026-10-16" }, { "label": "Nyne AI, data partnerships", "href": "https://nyne.ai/data-partnerships", "date": "2026-10-19" }, { "label": "Nyne AI, referrals", "href": "https://nyne.ai/referrals", "date": "2026-10-14" }, { "label": "Who Sells Data, buying programs register", "href": "https://whosellsdata.com/buying", "date": "2026-10-14" } ] }, { "slug": "ooak-data", "name": "Ooak Data", "kind": "Data partner", "url": "https://marindata.ai/buyers/ooak-data", "summary": "Ooak Data licenses company data from work tools for AI training and works with operating, winding-down and closed companies. It says its closed deals range from $20K to several hundred thousand dollars.", "goodFitFor": "Smaller companies, including closed ones, for whom a deal in the tens or low hundreds of thousands is worthwhile.", "watchFor": [ "Ooak's page said \"up to $1M\" on Oct 16, 2026. The Who Sells Data register recorded a $300K ceiling on Oct 14, 2026.", "Ooak receives the raw data before anonymizing it and commits by contract to delete raw archives within 30 days of delivery (Ooak, observed Oct 16, 2026)." ], "lastVerified": "2026-11-02", "termsPublished": 3, "notes": [], "terms": { "whatTheyBuy": { "label": "What they license", "value": "Company data from work tools, valued on years of history, headcount and how many tools connect and how deeply", "source": { "label": "Who Sells Data, buying programs register", "href": "https://whosellsdata.com/buying", "date": "2026-10-14" } }, "priceOrBands": { "label": "Price or bands", "value": "Headline \"up to $1M\"; closed deals from $20K to several hundred thousand dollars", "source": { "label": "Ooak Data, partner page", "href": "https://ooakdata.com/partner", "date": "2026-10-16" } }, "headcountFloor": { "label": "Headcount floor", "value": "Not published", "source": null }, "historyFloor": { "label": "Years of history", "value": "Not published", "source": null }, "geography": { "label": "Geography", "value": "Not published", "source": null }, "exclusivity": { "label": "Exclusivity", "value": "Not published", "source": null }, "paymentTiming": { "label": "Payment timing", "value": "Not published", "source": null }, "whoDeidentifies": { "label": "Who de-identifies", "value": "Ooak, with an in-house engine; only Ooak sees the raw data", "source": { "label": "Ooak Data, partner page", "href": "https://ooakdata.com/partner", "date": "2026-10-16" } }, "referralBounty": { "label": "Referral bounty", "value": "Not published", "source": null }, "publishedTermsPage": { "label": "Published terms page", "value": "Not published", "source": null } }, "sources": [ { "label": "Ooak Data, partner page", "href": "https://ooakdata.com/partner", "date": "2026-10-16" }, { "label": "Who Sells Data, buying programs register", "href": "https://whosellsdata.com/buying", "date": "2026-10-14" } ] }, { "slug": "opendatabay", "name": "Opendatabay", "kind": "Marketplace", "url": "https://marindata.ai/buyers/opendatabay", "summary": "Opendatabay is a licensed data marketplace open to registered organizations after identity checks. Listing is free, and sellers keep 70% to 95% of each sale, rising with transaction size.", "goodFitFor": "Organizations with packaged datasets that want to list them on a marketplace and accept non-exclusive sales.", "watchFor": [ "Commission also applies to off-platform deals with buyers Opendatabay introduced, for up to 24 months (Who Sells Data, Oct 14, 2026).", "The fee schedule recorded on Oct 16, 2026 a 12-month bar on exclusive licenses over delivered files. The schedule as of Oct 19, 2026 no longer states it (Opendatabay)." ], "lastVerified": "2026-11-02", "termsPublished": 4, "notes": [], "terms": { "whatTheyBuy": { "label": "What they license", "value": "Text, image, audio, video, code, agentic trajectories, 3D spatial, tabular, time-series, human feedback and synthetic data", "source": { "label": "Opendatabay, fees", "href": "https://www.opendatabay.com/legal/fees", "date": "2026-10-14" } }, "priceOrBands": { "label": "Price or bands", "value": "Sellers keep 70%–95% of each sale; listing is free", "source": { "label": "Opendatabay, fees", "href": "https://www.opendatabay.com/legal/fees", "date": "2026-10-14" } }, "headcountFloor": { "label": "Headcount floor", "value": "Registered businesses and organizations only; individuals cannot sell", "source": { "label": "Opendatabay, fees", "href": "https://www.opendatabay.com/legal/fees", "date": "2026-10-14" } }, "historyFloor": { "label": "Years of history", "value": "Not published", "source": null }, "geography": { "label": "Geography", "value": "Not published", "source": null }, "exclusivity": { "label": "Exclusivity", "value": "Not published", "source": null }, "paymentTiming": { "label": "Payment timing", "value": "Not published", "source": null }, "whoDeidentifies": { "label": "Who de-identifies", "value": "Not published", "source": null }, "referralBounty": { "label": "Referral bounty", "value": "Not published", "source": null }, "publishedTermsPage": { "label": "Published terms page", "value": "Fee schedule, last updated Oct 6, 2026", "source": { "label": "Opendatabay, fees", "href": "https://www.opendatabay.com/legal/fees", "date": "2026-10-14" } } }, "sources": [ { "label": "Opendatabay, fees", "href": "https://www.opendatabay.com/legal/fees", "date": "2026-10-14" }, { "label": "Opendatabay, fees (updated Oct 6, 2026)", "href": "https://www.opendatabay.com/legal/fees", "date": "2026-10-19" }, { "label": "Who Sells Data, buying programs register", "href": "https://whosellsdata.com/buying", "date": "2026-10-14" } ] }, { "slug": "pewter", "name": "Pewter", "kind": "Intermediary", "url": "https://marindata.ai/buyers/pewter", "summary": "Pewter is a deal intermediary in St. Pete Beach, Florida, that runs data-licensing sales for US firms with 20 or more employees. Its fee is paid by the licensing partner when a deal closes, and the owner pays nothing.", "goodFitFor": "US firms of 20 or more staff in logistics, construction, engineering, accounting and similar trades that want an intermediary at no cost to the owner.", "watchFor": [ "Pewter's fee is paid by the licensing partner, the buyer side of the deal (Pewter, observed Oct 16, 2026).", "Pewter's page says its deal figures are the publishers' own statements and have not been independently verified.", "Pewter states it is not a law firm or a registered broker-dealer (Pewter, observed Oct 16, 2026).", "Referral commission is paid only after Pewter is paid, which the page says can be months after the business is paid (Pewter, Nov 2, 2026)." ], "lastVerified": "2026-11-02", "termsPublished": 9, "notes": [], "terms": { "whatTheyBuy": { "label": "What they license", "value": "De-identified copies of email, shared drives, Slack, Jira and operating systems, such as rate quotes, shipment documents, CAD drawings, workpapers and SOPs", "source": { "label": "Pewter, homepage", "href": "https://pewterdata.com/", "date": "2026-10-16" } }, "priceOrBands": { "label": "Price or bands", "value": "Published closed deals of $110K–$1.2M (Pewter's own figures)", "source": { "label": "Pewter, homepage", "href": "https://pewterdata.com/", "date": "2026-10-16" } }, "headcountFloor": { "label": "Headcount floor", "value": "20+ full-time staff, now or at peak", "source": { "label": "Pewter, homepage", "href": "https://pewterdata.com/", "date": "2026-10-16" } }, "historyFloor": { "label": "Years of history", "value": "3+ years of records still on hand", "source": { "label": "Pewter, homepage", "href": "https://pewterdata.com/", "date": "2026-10-16" } }, "geography": { "label": "Geography", "value": "US-based, working mainly in English", "source": { "label": "Pewter, homepage", "href": "https://pewterdata.com/", "date": "2026-10-16" } }, "exclusivity": { "label": "Exclusivity", "value": "Negotiated per deal; several published deals were perpetual and exclusive", "source": { "label": "Pewter, homepage", "href": "https://pewterdata.com/", "date": "2026-10-16" } }, "paymentTiming": { "label": "Payment timing", "value": "Milestones in the written offer; published examples paid 5–9 days from bid", "source": { "label": "Pewter, homepage", "href": "https://pewterdata.com/", "date": "2026-10-16" } }, "whoDeidentifies": { "label": "Who de-identifies", "value": "Not published", "source": null }, "referralBounty": { "label": "Referral bounty", "value": "10% of the payment to the business, up to $100,000 per firm, paid once per business after Pewter is paid; a signed partner agreement sets the final terms", "source": { "label": "Pewter, referral partners", "href": "https://pewterdata.com/partners", "date": "2026-11-02" } }, "publishedTermsPage": { "label": "Published terms page", "value": "Referral partner page with commission terms", "source": { "label": "Pewter, referral partners", "href": "https://pewterdata.com/partners", "date": "2026-11-02" } } }, "sources": [ { "label": "Pewter, homepage", "href": "https://pewterdata.com/", "date": "2026-10-16" }, { "label": "Pewter, referral partners", "href": "https://pewterdata.com/partners", "date": "2026-11-02" } ] }, { "slug": "polyshares", "name": "Polyshares", "kind": "Data partner", "url": "https://marindata.ai/buyers/polyshares", "summary": "Polyshares licenses company operating records directly and is the counterparty on each deal. It publishes $100K to $2M+ per license and says it runs no marketplace, listing or auction.", "goodFitFor": "Companies in any sector with a long operating record that want one counterparty and an offer within two weeks.", "watchFor": [ "All figures and case studies are Polyshares' own; Who Sells Data found no press or funding coverage (Oct 14, 2026).", "The de-identification partner is not named (Polyshares, observed Oct 16, 2026)." ], "lastVerified": "2026-11-02", "termsPublished": 5, "notes": [], "terms": { "whatTheyBuy": { "label": "What they license", "value": "A company's operating record, under a defined, bounded license for a set term", "source": { "label": "Polyshares, homepage", "href": "https://polyshares.com/", "date": "2026-10-16" } }, "priceOrBands": { "label": "Price or bands", "value": "$100K–$2M+ per license", "source": { "label": "Polyshares, homepage", "href": "https://polyshares.com/", "date": "2026-10-16" } }, "headcountFloor": { "label": "Headcount floor", "value": "None stated; the length and texture of the record matter more than headcount", "source": { "label": "Polyshares, homepage", "href": "https://polyshares.com/", "date": "2026-10-16" } }, "historyFloor": { "label": "Years of history", "value": "Not published", "source": null }, "geography": { "label": "Geography", "value": "Not published", "source": null }, "exclusivity": { "label": "Exclusivity", "value": "Not published", "source": null }, "paymentTiming": { "label": "Payment timing", "value": "Net 30–60 once the data is approved, anonymized and delivered; offers within 1–2 weeks of intake", "source": { "label": "Polyshares, homepage", "href": "https://polyshares.com/", "date": "2026-10-16" } }, "whoDeidentifies": { "label": "Who de-identifies", "value": "An unnamed specialist PII partner, before Polyshares or any lab sees the record", "source": { "label": "Polyshares, homepage", "href": "https://polyshares.com/", "date": "2026-10-16" } }, "referralBounty": { "label": "Referral bounty", "value": "Not published", "source": null }, "publishedTermsPage": { "label": "Published terms page", "value": "Not published", "source": null } }, "sources": [ { "label": "Polyshares, homepage", "href": "https://polyshares.com/", "date": "2026-10-16" }, { "label": "Who Sells Data, buying programs register", "href": "https://whosellsdata.com/buying", "date": "2026-10-14" } ] }, { "slug": "pre-dev", "name": "pre.dev", "kind": "Marketplace", "url": "https://marindata.ai/buyers/pre-dev", "summary": "pre.dev, a New York coding-agent company, licenses private codebases with their git history to AI labs at a published per-token rate. Its licenses are non-exclusive by default, so the same code can be paid for more than once.", "goodFitFor": "Software owners who hold clean title to their code and want a fixed per-token price and the right to license the same code again.", "watchFor": [ "Payment depends on pre.dev first receiving the buyer's funds (Who Sells Data, Oct 14, 2026).", "Company and client names in paths, comments and documents are left as written after scrubbing (pre.dev, observed Oct 17, 2026).", "The owner's indemnity is uncapped, while pre.dev's liability is capped at what it paid the owner in the prior 12 months (pre.dev Codebase Agreement, observed Oct 17, 2026)." ], "lastVerified": "2026-11-02", "termsPublished": 8, "notes": [ { "text": "The homepage example prices a repo with 20 million source tokens and 100 million history tokens at $6,000.", "source": { "label": "pre.dev, code licensing", "href": "https://pre.dev/#code-licensing", "date": "2026-10-17" } }, { "text": "pre.dev's estimator counts tokens as bytes divided by four, caps history at 100MB per path and skips vendored, minified and generated files.", "source": { "label": "pre.dev, token estimator script", "href": "https://api.pre.dev/api/code-licensing/estimate.mjs", "date": "2026-10-17" } }, { "text": "One signature authorizes all later standard non-exclusive licenses, so the owner does not approve each deal. An interested AI lab reviews the repo's metadata, never the code, for 4 to 12 weeks, then makes an offer or passes.", "source": { "label": "pre.dev, license your codebases", "href": "https://pre.dev/code-licensing", "date": "2026-10-17" } }, { "text": "The standing rates are the payout itself, with no platform commission taken from them. The owner is not told what the buyer paid or who the buyer is.", "source": { "label": "pre.dev Codebase Agreement (PDF)", "href": "https://api.pre.dev/api/code-licensing/agreement.pdf", "date": "2026-10-17" } }, { "text": "Buyer licenses are perpetual and irrevocable once they close, and buyers may use the code to train and fine-tune models. pre.dev may also take its own perpetual non-exclusive license at the standing rates.", "source": { "label": "pre.dev Codebase Agreement (PDF)", "href": "https://api.pre.dev/api/code-licensing/agreement.pdf", "date": "2026-10-17" } }, { "text": "Withdrawing takes 30 days' written notice. Existing licenses continue, deals already under way have a 90-day tail, and backups and pre.dev's own licensed copies are not destroyed.", "source": { "label": "pre.dev Codebase Agreement (PDF)", "href": "https://api.pre.dev/api/code-licensing/agreement.pdf", "date": "2026-10-17" } }, { "text": "The owner warrants a full chain of title from every founder, employee and contractor, that no client holds rights in the code, and that every open-source component and its license is disclosed. Agencies licensing client work should check this first.", "source": { "label": "pre.dev Codebase Agreement (PDF)", "href": "https://api.pre.dev/api/code-licensing/agreement.pdf", "date": "2026-10-17" } }, { "text": "The agreement bars dealing directly with a buyer pre.dev introduced, during the term and for 12 months after.", "source": { "label": "pre.dev Codebase Agreement (PDF)", "href": "https://api.pre.dev/api/code-licensing/agreement.pdf", "date": "2026-10-17" } } ], "terms": { "whatTheyBuy": { "label": "What they license", "value": "Private codebases with their git history; owners connect a Git account and choose repositories. Public repos, forks, archived repos and published packages are excluded", "source": { "label": "pre.dev, llms.txt site summary", "href": "https://pre.dev/llms.txt", "date": "2026-10-17" } }, "priceOrBands": { "label": "Price or bands", "value": "$100 per million source tokens plus $40 per million git-history tokens, per non-exclusive license; pre.dev can change the rates on 30 days' notice", "source": { "label": "pre.dev Codebase Agreement (PDF)", "href": "https://api.pre.dev/api/code-licensing/agreement.pdf", "date": "2026-10-17" } }, "headcountFloor": { "label": "Headcount floor", "value": "None; individual developers and companies or agencies can both sign up", "source": { "label": "pre.dev, license your codebases", "href": "https://pre.dev/code-licensing", "date": "2026-10-17" } }, "historyFloor": { "label": "Years of history", "value": "Not published", "source": null }, "geography": { "label": "Geography", "value": "Contract with Pre Dev Corp under Delaware law; paid in US dollars by ACH or wire, with W-8BEN forms accepted for non-US owners", "source": { "label": "pre.dev Codebase Agreement (PDF)", "href": "https://api.pre.dev/api/code-licensing/agreement.pdf", "date": "2026-10-17" } }, "exclusivity": { "label": "Exclusivity", "value": "Non-exclusive by default; an exclusive license needs a separate confirmation signed by both sides, which sets the price", "source": { "label": "pre.dev Codebase Agreement (PDF)", "href": "https://api.pre.dev/api/code-licensing/agreement.pdf", "date": "2026-10-17" } }, "paymentTiming": { "label": "Payment timing", "value": "Within 15 days of the buyer's payment clearing; pre.dev says buyers commonly use 60-day payment terms", "source": { "label": "pre.dev, license your codebases", "href": "https://pre.dev/code-licensing", "date": "2026-10-17" } }, "whoDeidentifies": { "label": "Who de-identifies", "value": "pre.dev scrubs secrets and personal data from its copy and maps commit authors to pseudonyms", "source": { "label": "pre.dev, llms.txt site summary", "href": "https://pre.dev/llms.txt", "date": "2026-10-17" } }, "referralBounty": { "label": "Referral bounty", "value": "Not published", "source": null }, "publishedTermsPage": { "label": "Published terms page", "value": "Yes: the pre.dev Codebase Agreement, a PDF linked from the licensing FAQ", "source": { "label": "pre.dev Codebase Agreement (PDF)", "href": "https://api.pre.dev/api/code-licensing/agreement.pdf", "date": "2026-10-17" } } }, "sources": [ { "label": "pre.dev, code licensing", "href": "https://pre.dev/#code-licensing", "date": "2026-10-17" }, { "label": "pre.dev, license your codebases", "href": "https://pre.dev/code-licensing", "date": "2026-10-17" }, { "label": "pre.dev Codebase Agreement (PDF)", "href": "https://api.pre.dev/api/code-licensing/agreement.pdf", "date": "2026-10-17" }, { "label": "pre.dev, llms.txt site summary", "href": "https://pre.dev/llms.txt", "date": "2026-10-17" }, { "label": "pre.dev, token estimator script", "href": "https://api.pre.dev/api/code-licensing/estimate.mjs", "date": "2026-10-17" }, { "label": "Who Sells Data, buying programs register", "href": "https://whosellsdata.com/buying", "date": "2026-10-14" } ] }, { "slug": "replay", "name": "Replay", "kind": "Data partner", "url": "https://marindata.ai/buyers/replay", "summary": "Replay licenses de-identified copies of company operating records to frontier AI and data labs. It pays an upfront fee plus a revenue share on every license.", "goodFitFor": "Companies of 20 or more staff with years of records who want an upfront payment and a share of later licenses.", "watchFor": [ "Replay calls headcount a proxy; years in operation, industry, apps and data volume set the final price (Replay, observed Oct 19, 2026).", "The referral fee is one-time on the upfront fee; referrers get no share of later revenue, and a company already in Replay's pipeline does not count (Replay, Oct 26, 2026).", "The revenue-share percentage is not published (Who Sells Data, Oct 16, 2026)." ], "lastVerified": "2026-11-02", "termsPublished": 8, "notes": [ { "text": "Replay says records are sold only to frontier AI and data labs, never to foreign adversaries, and never resold through brokers. It estimates the process takes about 120 minutes of the company's time.", "source": { "label": "Replay, homepage", "href": "https://www.replay.ai/", "date": "2026-10-19" } } ], "terms": { "whatTheyBuy": { "label": "What they license", "value": "Code and commits, decision threads, financial records, support history, product logs, presentations, CRM data, sales conversations, internal docs and analytics", "source": { "label": "Replay, homepage", "href": "https://www.replay.ai/", "date": "2026-10-19" } }, "priceOrBands": { "label": "Price or bands", "value": "$10K–$100K for 20–50 employees, $100K–$1M for 50–250, $1M+ for 250+", "source": { "label": "Replay, homepage", "href": "https://www.replay.ai/", "date": "2026-10-19" } }, "headcountFloor": { "label": "Headcount floor", "value": "20 employees (lowest published band)", "source": { "label": "Replay, homepage", "href": "https://www.replay.ai/", "date": "2026-10-19" } }, "historyFloor": { "label": "Years of history", "value": "Not published", "source": null }, "geography": { "label": "Geography", "value": "English-language operations, US preferred", "source": { "label": "Who Sells Data, Replay entry", "href": "https://whosellsdata.com/buying", "date": "2026-10-16" } }, "exclusivity": { "label": "Exclusivity", "value": "Not published", "source": null }, "paymentTiming": { "label": "Payment timing", "value": "Paid up front, plus a revenue share of every license", "source": { "label": "Replay, homepage", "href": "https://www.replay.ai/", "date": "2026-10-19" } }, "whoDeidentifies": { "label": "Who de-identifies", "value": "Replay, during export", "source": { "label": "Replay, homepage", "href": "https://www.replay.ai/", "date": "2026-10-19" } }, "referralBounty": { "label": "Referral bounty", "value": "$10,000 or 10% of the upfront fee, whichever is greater, paid the day the company's upfront fee lands; one-time per company, no cap, and paid by Replay on top of the company's offer", "source": { "label": "Replay, referrals", "href": "https://www.replay.ai/referrals", "date": "2026-10-26" } }, "publishedTermsPage": { "label": "Published terms page", "value": "Terms of use at /terms", "source": { "label": "Replay, homepage", "href": "https://www.replay.ai/", "date": "2026-10-19" } } }, "sources": [ { "label": "Replay, homepage", "href": "https://www.replay.ai/", "date": "2026-10-19" }, { "label": "Replay, referrals", "href": "https://www.replay.ai/referrals", "date": "2026-10-26" }, { "label": "Who Sells Data, Replay entry", "href": "https://whosellsdata.com/buying", "date": "2026-10-16" } ] }, { "slug": "scale-ai", "name": "Scale AI", "kind": "Data partner", "url": "https://marindata.ai/buyers/scale-ai", "summary": "Scale AI is a large AI training data company. Its Data Partnerships program licenses de-identified company records, anonymized inside the company's own environment before extraction.", "goodFitFor": "Companies of 40 or more staff with established data systems, a clear internal owner and interest in recurring deliveries.", "watchFor": [ "Scale labels the $10K–$1M+ range illustrative (Scale, observed Oct 16, 2026).", "No separate referral terms are published (Who Sells Data, Oct 14, 2026)." ], "lastVerified": "2026-11-02", "termsPublished": 5, "notes": [], "terms": { "whatTheyBuy": { "label": "What they license", "value": "Structured, de-identified records from data warehouses, chat, email and calendar, documents, work and knowledge tools, and business platforms", "source": { "label": "Scale AI, data partnership", "href": "https://scale.com/data-partnership", "date": "2026-10-16" } }, "priceOrBands": { "label": "Price or bands", "value": "Illustrative $10K–$1M+ per partnership, scaling with delivery cadence", "source": { "label": "Scale AI, data partnership", "href": "https://scale.com/data-partnership", "date": "2026-10-16" } }, "headcountFloor": { "label": "Headcount floor", "value": "40+ employees preferred; assessed case by case", "source": { "label": "Scale AI, data partnership", "href": "https://scale.com/data-partnership", "date": "2026-10-16" } }, "historyFloor": { "label": "Years of history", "value": "Not published", "source": null }, "geography": { "label": "Geography", "value": "Not published", "source": null }, "exclusivity": { "label": "Exclusivity", "value": "Not published", "source": null }, "paymentTiming": { "label": "Payment timing", "value": "Not published", "source": null }, "whoDeidentifies": { "label": "Who de-identifies", "value": "Scale, inside the company's environment before any data is extracted", "source": { "label": "Scale AI, data partnership", "href": "https://scale.com/data-partnership", "date": "2026-10-16" } }, "referralBounty": { "label": "Referral bounty", "value": "Up to $50K per referred company, paid only if Scale signs a partnership; one payout per deal", "source": { "label": "Scale AI, data partnership", "href": "https://scale.com/data-partnership", "date": "2026-10-16" } }, "publishedTermsPage": { "label": "Published terms page", "value": "Not published", "source": null } }, "sources": [ { "label": "Scale AI, data partnership", "href": "https://scale.com/data-partnership", "date": "2026-10-16" }, { "label": "Who Sells Data, buying programs register", "href": "https://whosellsdata.com/buying", "date": "2026-10-14" } ] }, { "slug": "simpleclosure-assethub", "name": "SimpleClosure AssetHub", "kind": "Wind-down platform", "url": "https://marindata.ai/buyers/simpleclosure-assethub", "summary": "SimpleClosure runs startup shutdowns, and its AssetHub brokers the sale or license of source code and workspace data to AI labs and other buyers. It says it has supported 350+ companies and is on track to return $10M+ to startups by the end of 2026.", "goodFitFor": "Venture-backed startups that are shutting down, especially those already using SimpleClosure for the dissolution.", "watchFor": [ "SimpleClosure takes a percentage on completed sales or licenses and does not publish the rate. AssetHub is free for its dissolution customers (Who Sells Data, Oct 14, 2026).", "Its page does not address consent from employees whose messages are in workspace data (Who Sells Data, Oct 14, 2026)." ], "lastVerified": "2026-11-02", "termsPublished": 3, "notes": [], "terms": { "whatTheyBuy": { "label": "What they license", "value": "Source code and repositories, plus workspace data such as communications, documents, designs, tickets and workflows", "source": { "label": "SimpleClosure, AssetHub", "href": "https://simpleclosure.com/product/assethub/", "date": "2026-10-16" } }, "priceOrBands": { "label": "Price or bands", "value": "Past deals ran from tens to hundreds of thousands of dollars", "source": { "label": "Who Sells Data, buying programs register", "href": "https://whosellsdata.com/buying", "date": "2026-10-14" } }, "headcountFloor": { "label": "Headcount floor", "value": "Not published", "source": null }, "historyFloor": { "label": "Years of history", "value": "Not published", "source": null }, "geography": { "label": "Geography", "value": "Not published", "source": null }, "exclusivity": { "label": "Exclusivity", "value": "Not published", "source": null }, "paymentTiming": { "label": "Payment timing", "value": "Not published", "source": null }, "whoDeidentifies": { "label": "Who de-identifies", "value": "SimpleClosure, through a multi-pass process; sensitive employee and health records are not sold", "source": { "label": "SimpleClosure AssetHub on track to return $10M+ (release via Webull)", "href": "https://www.webull.com/news/15515143679517696", "date": "2026-09-03" } }, "referralBounty": { "label": "Referral bounty", "value": "Not published", "source": null }, "publishedTermsPage": { "label": "Published terms page", "value": "Not published", "source": null } }, "sources": [ { "label": "SimpleClosure, AssetHub", "href": "https://simpleclosure.com/product/assethub/", "date": "2026-10-16" }, { "label": "SimpleClosure AssetHub on track to return $10M+ (release via Webull)", "href": "https://www.webull.com/news/15515143679517696", "date": "2026-09-03" }, { "label": "Who Sells Data, buying programs register", "href": "https://whosellsdata.com/buying", "date": "2026-10-14" }, { "label": "MIT Technology Review, AI labs are buying data from dead startups", "href": "https://technologyreview.com/2026/10/13/1145755/ai-labs-buying-data-from-dead-startups-to-train-models", "date": "2026-10-13" } ] }, { "slug": "superannotate", "name": "SuperAnnotate", "kind": "Licensing platform", "url": "https://marindata.ai/buyers/superannotate", "summary": "SuperAnnotate prepares companies' operational records and licenses them to frontier AI labs. Rights go for a set period to buyers the company approves, with a payment on each relicense.", "goodFitFor": "Firms of any size with specialized records that want to approve every buyer and earn on relicensing.", "watchFor": [ "The length of the license period is not stated (SuperAnnotate, observed Oct 16, 2026).", "SuperAnnotate does not publish its own fee or revenue share (observed Oct 16, 2026)." ], "lastVerified": "2026-11-02", "termsPublished": 6, "notes": [], "terms": { "whatTheyBuy": { "label": "What they license", "value": "Business records built up over time that are specific to the field and that the company has the rights to license", "source": { "label": "SuperAnnotate, data partnerships", "href": "https://www.superannotate.com/data-partnerships", "date": "2026-10-16" } }, "priceOrBands": { "label": "Price or bands", "value": "Indicative $100K–$1M+, depending on rarity, depth and exclusivity", "source": { "label": "SuperAnnotate, data partnerships", "href": "https://www.superannotate.com/data-partnerships", "date": "2026-10-16" } }, "headcountFloor": { "label": "Headcount floor", "value": "None; any company size, industry or location", "source": { "label": "SuperAnnotate, data partnerships", "href": "https://www.superannotate.com/data-partnerships", "date": "2026-10-16" } }, "historyFloor": { "label": "Years of history", "value": "Not published", "source": null }, "geography": { "label": "Geography", "value": "Any location", "source": { "label": "SuperAnnotate, data partnerships", "href": "https://www.superannotate.com/data-partnerships", "date": "2026-10-16" } }, "exclusivity": { "label": "Exclusivity", "value": "Priced into the valuation; the company approves each buyer and can block competitors", "source": { "label": "SuperAnnotate, data partnerships", "href": "https://www.superannotate.com/data-partnerships", "date": "2026-10-16" } }, "paymentTiming": { "label": "Payment timing", "value": "On delivery, and again on each relicense", "source": { "label": "SuperAnnotate, data partnerships", "href": "https://www.superannotate.com/data-partnerships", "date": "2026-10-16" } }, "whoDeidentifies": { "label": "Who de-identifies", "value": "Not published", "source": null }, "referralBounty": { "label": "Referral bounty", "value": "Not published", "source": null }, "publishedTermsPage": { "label": "Published terms page", "value": "Not published", "source": null } }, "sources": [ { "label": "SuperAnnotate, data partnerships", "href": "https://www.superannotate.com/data-partnerships", "date": "2026-10-16" }, { "label": "Who Sells Data, buying programs register", "href": "https://whosellsdata.com/buying", "date": "2026-10-14" } ] }, { "slug": "surge-ai", "name": "Surge AI", "kind": "Data partner", "url": "https://marindata.ai/buyers/surge-ai", "summary": "Surge AI supplies training data to AI labs. MIT Technology Review reported on Oct 13, 2026 that it buys company archives to build reinforcement-learning \"gyms\" for training AI agents.", "goodFitFor": "Companies that Surge approaches directly. We found no public seller intake or terms page as of Oct 17, 2026.", "watchFor": [ "Surge publishes no seller terms, so every figure comes from direct negotiation (observed Oct 17, 2026).", "Surge is not listed on the Who Sells Data register of buying programs (Who Sells Data, observed Oct 17, 2026)." ], "lastVerified": "2026-11-02", "termsPublished": 1, "notes": [ { "text": "Per MIT Technology Review, Surge builds reinforcement-learning \"gyms\" where its clients' AI agents practice tasks such as debugging IT architecture and drafting appellate briefs, and it buys archives from failed companies to build them. Nick Heiner leads the work.", "source": { "label": "MIT Technology Review, AI labs are buying data from dead startups", "href": "https://technologyreview.com/2026/10/13/1145755/ai-labs-buying-data-from-dead-startups-to-train-models", "date": "2026-10-13" } }, { "text": "Heiner told the magazine he looks for archives that show a range of situations, \"from disagreements to tricky judgment calls\". He said one company's detailed Slack history can be worth more than another company's Slack, email and calendars combined.", "source": { "label": "MIT Technology Review, AI labs are buying data from dead startups", "href": "https://technologyreview.com/2026/10/13/1145755/ai-labs-buying-data-from-dead-startups-to-train-models", "date": "2026-10-13" } }, { "text": "Surge's website has no page for companies selling data. None of the 94 URLs in its sitemap is a seller or data-partnership page, and its contact points are for buyers requesting datasets.", "source": { "label": "Surge AI, sitemap", "href": "https://surgehq.ai/sitemap.xml", "date": "2026-10-26" } }, { "text": "For the datasets it sells, Surge offers labs a program under which they pay only if the data improves their model.", "source": { "label": "Surge AI, off-the-shelf frontier data", "href": "https://surgehq.ai/ots", "date": "2026-10-17" } }, { "text": "Surge's own writing on RL environments describes ones it builds, such as CoreCraft Inc., a fictional PC-parts retailer. The archive buying is known from press reporting only.", "source": { "label": "Surge AI, RL environments and the hierarchy of agentic capabilities", "href": "https://surgehq.ai/blog/rl-envs-real-world", "date": "2026-02-05" } }, { "text": "Surge says it works \"in partnership with companies like Anthropic, Google, Microsoft, and Meta\".", "source": { "label": "Surge AI, Strategic Projects Lead job posting", "href": "https://surgehq.ai/careers/strategic-projects-lead", "date": "2026-10-17" } }, { "text": "Forbes reports that Surge had $1.2 billion of revenue in 2024, has raised no outside funding, and that founder Edwin Chen owns about 75% of it.", "source": { "label": "Forbes, Edwin Chen profile", "href": "https://www.forbes.com/profile/edwin-chen/", "date": "2026-09-15" } }, { "text": "A worker misclassification class action was filed against Surge in San Francisco Superior Court in May 2025. It concerns the people who label data for Surge.", "source": { "label": "Bloomberg Law, Surge AI hit with worker misclassification suit", "href": "https://news.bloomberglaw.com/financial-accounting/ai-training-firm-surge-ai-hit-with-worker-misclassification-suit", "date": "2025-05-21" } } ], "terms": { "whatTheyBuy": { "label": "What they license", "value": "Company archives used to build reinforcement-learning training environments", "source": { "label": "MIT Technology Review, AI labs are buying data from dead startups", "href": "https://technologyreview.com/2026/10/13/1145755/ai-labs-buying-data-from-dead-startups-to-train-models", "date": "2026-10-13" } }, "priceOrBands": { "label": "Price or bands", "value": "Not published", "source": null }, "headcountFloor": { "label": "Headcount floor", "value": "Not published", "source": null }, "historyFloor": { "label": "Years of history", "value": "Not published", "source": null }, "geography": { "label": "Geography", "value": "Not published", "source": null }, "exclusivity": { "label": "Exclusivity", "value": "Not published", "source": null }, "paymentTiming": { "label": "Payment timing", "value": "Not published", "source": null }, "whoDeidentifies": { "label": "Who de-identifies", "value": "Not published", "source": null }, "referralBounty": { "label": "Referral bounty", "value": "Not published", "source": null }, "publishedTermsPage": { "label": "Published terms page", "value": "Not published", "source": null } }, "sources": [ { "label": "MIT Technology Review, AI labs are buying data from dead startups", "href": "https://technologyreview.com/2026/10/13/1145755/ai-labs-buying-data-from-dead-startups-to-train-models", "date": "2026-10-13" }, { "label": "Surge AI, sitemap", "href": "https://surgehq.ai/sitemap.xml", "date": "2026-10-26" }, { "label": "Surge AI, off-the-shelf frontier data", "href": "https://surgehq.ai/ots", "date": "2026-10-17" }, { "label": "Surge AI, RL environments and the hierarchy of agentic capabilities", "href": "https://surgehq.ai/blog/rl-envs-real-world", "date": "2026-02-05" }, { "label": "Surge AI, Strategic Projects Lead job posting", "href": "https://surgehq.ai/careers/strategic-projects-lead", "date": "2026-10-17" }, { "label": "Forbes, Edwin Chen profile", "href": "https://www.forbes.com/profile/edwin-chen/", "date": "2026-09-15" }, { "label": "Bloomberg Law, Surge AI hit with worker misclassification suit", "href": "https://news.bloomberglaw.com/financial-accounting/ai-training-firm-surge-ai-hit-with-worker-misclassification-suit", "date": "2025-05-21" }, { "label": "Who Sells Data, buying programs register (updated Oct 15, 2026)", "href": "https://whosellsdata.com/buying", "date": "2026-10-17" } ] }, { "slug": "the-general-data-company", "name": "The General Data Company", "kind": "Data partner", "url": "https://marindata.ai/buyers/the-general-data-company", "summary": "The General Data Company licenses the operating records of industrial companies to AI labs and to manufacturers training their own models. It pays either a one-off price or a share of what it earns from licensing.", "goodFitFor": "Industrial and engineering firms, including ones no longer trading, that hold the rights to their project, maintenance and CAD records.", "watchFor": [ "Customer or supplier records qualify only where the company holds the rights and no confidentiality agreement stands in the way (Who Sells Data, Oct 14, 2026)." ], "lastVerified": "2026-11-02", "termsPublished": 4, "notes": [], "terms": { "whatTheyBuy": { "label": "What they license", "value": "Completed projects, correspondence, orders, technical documentation, complaints and maintenance records, ticket and CRM histories, drawings and CAD models", "source": { "label": "The General Data Company, data partnerships", "href": "https://thegeneraldata.com/data-partnerships/", "date": "2026-10-14" } }, "priceOrBands": { "label": "Price or bands", "value": "Not published", "source": null }, "headcountFloor": { "label": "Headcount floor", "value": "Not published", "source": null }, "historyFloor": { "label": "Years of history", "value": "Not published", "source": null }, "geography": { "label": "Geography", "value": "Not published", "source": null }, "exclusivity": { "label": "Exclusivity", "value": "Not published", "source": null }, "paymentTiming": { "label": "Payment timing", "value": "Follows milestones agreed in writing", "source": { "label": "Who Sells Data, buying programs register", "href": "https://whosellsdata.com/buying", "date": "2026-10-14" } }, "whoDeidentifies": { "label": "Who de-identifies", "value": "The General Data Company removes personal data or replaces it with fictional values first", "source": { "label": "Who Sells Data, buying programs register", "href": "https://whosellsdata.com/buying", "date": "2026-10-14" } }, "referralBounty": { "label": "Referral bounty", "value": "Up to $50K per introduction, set in a separate written referral agreement and paid on a signed partnership", "source": { "label": "The General Data Company, data partnerships", "href": "https://thegeneraldata.com/data-partnerships/", "date": "2026-10-14" } }, "publishedTermsPage": { "label": "Published terms page", "value": "Not published", "source": null } }, "sources": [ { "label": "The General Data Company, data partnerships", "href": "https://thegeneraldata.com/data-partnerships/", "date": "2026-10-14" }, { "label": "Who Sells Data, buying programs register", "href": "https://whosellsdata.com/buying", "date": "2026-10-14" } ] }, { "slug": "troveo", "name": "Troveo AI", "kind": "Licensing platform", "url": "https://marindata.ai/buyers/troveo", "summary": "Troveo AI licenses content and company data to AI labs on the owner's behalf and takes a share of each sale. It reports more than $20M paid to licensors and 40+ active buyers.", "goodFitFor": "Owners who want their data licensed to several buyers over time and accept a revenue share in place of a fixed price.", "watchFor": [ "Payout totals and example figures on Troveo's page are its own (Who Sells Data, Oct 14, 2026).", "Buyer names are disclosed only after an agreement is signed (Troveo, observed Oct 16, 2026)." ], "lastVerified": "2026-11-02", "termsPublished": 3, "notes": [], "terms": { "whatTheyBuy": { "label": "What they license", "value": "Company data exports from chat, email, documents, CRM and support, finance and operations, and code repositories; also video, audio and text content", "source": { "label": "Troveo, business data", "href": "https://www.troveo.ai/business-data", "date": "2026-10-16" } }, "priceOrBands": { "label": "Price or bands", "value": "Full-company deals typically start at six figures; revenue share on every sale, with no percentage published", "source": { "label": "Troveo, business data", "href": "https://www.troveo.ai/business-data", "date": "2026-10-16" } }, "headcountFloor": { "label": "Headcount floor", "value": "Not published", "source": null }, "historyFloor": { "label": "Years of history", "value": "Not published", "source": null }, "geography": { "label": "Geography", "value": "Not published", "source": null }, "exclusivity": { "label": "Exclusivity", "value": "Not published", "source": null }, "paymentTiming": { "label": "Payment timing", "value": "Not published", "source": null }, "whoDeidentifies": { "label": "Who de-identifies", "value": "Troveo, to what it calls a documented standard", "source": { "label": "Troveo, business data", "href": "https://www.troveo.ai/business-data", "date": "2026-10-16" } }, "referralBounty": { "label": "Referral bounty", "value": "Not published", "source": null }, "publishedTermsPage": { "label": "Published terms page", "value": "Not published", "source": null } }, "sources": [ { "label": "Troveo, business data", "href": "https://www.troveo.ai/business-data", "date": "2026-10-16" }, { "label": "Business Wire, Troveo announces $20M in payouts", "href": "https://www.businesswire.com/news/home/20260428319383/en/Troveo-Accelerates-AI-Model-Development-Expands-AI-Training-Data-Platform-to-Five-New-Categories-Announces-%2420-Million-in-Payouts", "date": "2026-04-28" }, { "label": "Who Sells Data, buying programs register", "href": "https://whosellsdata.com/buying", "date": "2026-10-14" }, { "label": "aitrainer.work, how companies get paid for operational data", "href": "https://aitrainer.work/enterprise/guides/how-companies-get-paid-for-operational-data", "date": "2026-10-04" } ] }, { "slug": "turing-project-lazarus", "name": "Turing (Project Lazarus)", "kind": "Data partner", "url": "https://marindata.ai/buyers/turing-project-lazarus", "summary": "Turing is an AI research and data company. Its Project Lazarus program buys defined rights to a company's operating record, from communications and tickets to source code, for a single payment.", "goodFitFor": "Operating, winding-down or closed companies with years of complete records across code and business tools.", "watchFor": [ "Code is not bought on its own (Who Sells Data, Oct 14, 2026).", "HUD, a competing marketplace, reports that Turing pays $10K–$100K for legacy code and up to $15K per modern repository (HUD, Jun 24, 2026). Turing's own page gives only the up-to-$1M figure for a full record.", "Turing publishes one firm timeline: an intake call within one business day and a sample in about a week, with the price after sampling (Turing, observed Oct 16, 2026)." ], "lastVerified": "2026-11-02", "termsPublished": 5, "notes": [], "terms": { "whatTheyBuy": { "label": "What they license", "value": "A company's operating record: communication threads, source code and reviews, tickets, wikis, runbooks, schemas and the reasoning behind decisions", "source": { "label": "Turing, Project Lazarus", "href": "https://lazarus.turing.com/", "date": "2026-10-16" } }, "priceOrBands": { "label": "Price or bands", "value": "A single payment of up to $1M, confirmed after sampling", "source": { "label": "Turing, Project Lazarus", "href": "https://lazarus.turing.com/", "date": "2026-10-16" } }, "headcountFloor": { "label": "Headcount floor", "value": "Not published", "source": null }, "historyFloor": { "label": "Years of history", "value": "Not published", "source": null }, "geography": { "label": "Geography", "value": "Not published", "source": null }, "exclusivity": { "label": "Exclusivity", "value": "Not published", "source": null }, "paymentTiming": { "label": "Payment timing", "value": "Not published", "source": null }, "whoDeidentifies": { "label": "Who de-identifies", "value": "Turing's Lazarus team removes personal data and secrets, working from a copy in an isolated environment", "source": { "label": "Turing, Project Lazarus", "href": "https://lazarus.turing.com/", "date": "2026-10-16" } }, "referralBounty": { "label": "Referral bounty", "value": "Up to $10K per referred company", "source": { "label": "Turing, Project Lazarus", "href": "https://lazarus.turing.com/", "date": "2026-10-16" } }, "publishedTermsPage": { "label": "Published terms page", "value": "Program page only; no referral terms page", "source": { "label": "Who Sells Data, buying programs register", "href": "https://whosellsdata.com/buying", "date": "2026-10-14" } } }, "sources": [ { "label": "Turing, Project Lazarus", "href": "https://lazarus.turing.com/", "date": "2026-10-16" }, { "label": "Who Sells Data, buying programs register", "href": "https://whosellsdata.com/buying", "date": "2026-10-14" }, { "label": "HUD, best platforms for selling a codebase", "href": "https://www.hud.ai/resources/best-platforms-selling-codebase", "date": "2026-06-24" }, { "label": "Let's Data Science, Turing buys failed-startup codebases", "href": "https://letsdatascience.com/news/turing-buys-failed-startup-codebases-for-training-7bb0d84f", "date": "2025-12-18" }, { "label": "aitrainer.work, how companies get paid for operational data", "href": "https://aitrainer.work/enterprise/guides/how-companies-get-paid-for-operational-data", "date": "2026-10-04" } ] }, { "slug": "verb", "name": "Verb", "kind": "Marketplace", "url": "https://marindata.ai/buyers/verb", "summary": "Verb is a data exchange where individuals sell their own activity, transaction and email data, and its CEO expects AI labs to be the largest buyers. It has no route for a company to sell its operational records.", "goodFitFor": "Individuals. Verb is not a program for companies licensing operational records.", "watchFor": [ "Verb says it excludes passwords, text messages and health data from analysis (Axios, Aug 6, 2026).", "Verb discloses no fee or commission on sales (Verb, observed Oct 17, 2026)." ], "lastVerified": "2026-10-17", "termsPublished": 6, "notes": [ { "text": "Buyers query the seller pool in aggregate, and nothing is shared unless a buyer meets the seller's reserve price. Sellers can block individual buyers.", "source": { "label": "Verb, homepage and seller dashboard", "href": "https://verb-data.com", "date": "2026-10-17" } }, { "text": "Verb grew out of Verb.AI, which Generation Lab, a youth polling company, launched in May 2025. Cyrus Beschloss is CEO.", "source": { "label": "Axios, Generation Lab launches Verb.AI", "href": "https://www.axios.com/2025/05/20/gen-z-data-privacy-generation-lab-verb-ai", "date": "2025-05-20" } }, { "text": "Companies appear on Verb only as buyers. Its only email route is a person forwarding their own mail, so it cannot carry a company archive (checked Oct 17, 2026).", "source": { "label": "Verb, homepage and seller dashboard", "href": "https://verb-data.com", "date": "2026-10-17" } } ], "terms": { "whatTheyBuy": { "label": "What they license", "value": "Individuals' phone activity (VPN tracker), bank and card transactions via Plaid, forwarded personal email, social media exports and AI assistant logs", "source": { "label": "Verb, homepage and seller dashboard", "href": "https://verb-data.com", "date": "2026-10-17" } }, "priceOrBands": { "label": "Price or bands", "value": "Each seller sets a reserve price in dollars per buyer per month; the tracker alone pays up to $30 a month", "source": { "label": "Verb, homepage and seller dashboard", "href": "https://verb-data.com", "date": "2026-10-17" } }, "headcountFloor": { "label": "Headcount floor", "value": "Individuals aged 18 or older; no seller route for companies", "source": { "label": "Verb, homepage and seller dashboard", "href": "https://verb-data.com", "date": "2026-10-17" } }, "historyFloor": { "label": "Years of history", "value": "Not published", "source": null }, "geography": { "label": "Geography", "value": "Operated from the US; payouts by US bank transfer (ACH)", "source": { "label": "Verb, terms of service", "href": "https://verb-data.com/terms-of-service", "date": "2026-10-17" } }, "exclusivity": { "label": "Exclusivity", "value": "Not published", "source": null }, "paymentTiming": { "label": "Payment timing", "value": "Credits pend for 7 days; ACH payout once $15 is earned, arriving in 1–3 business days", "source": { "label": "Verb, homepage and seller dashboard", "href": "https://verb-data.com", "date": "2026-10-17" } }, "whoDeidentifies": { "label": "Who de-identifies", "value": "Not published", "source": null }, "referralBounty": { "label": "Referral bounty", "value": "Not published", "source": null }, "publishedTermsPage": { "label": "Published terms page", "value": "Terms of service and privacy policy", "source": { "label": "Verb, terms of service", "href": "https://verb-data.com/terms-of-service", "date": "2026-10-17" } } }, "sources": [ { "label": "Verb, homepage and seller dashboard", "href": "https://verb-data.com", "date": "2026-10-17" }, { "label": "Verb, terms of service", "href": "https://verb-data.com/terms-of-service", "date": "2026-10-17" }, { "label": "Axios, exclusive on Verb", "href": "https://www.axios.com/2026/08/06/exclusive-verb-lets-you-sell-your-data-to-ai-labs-for-training", "date": "2026-08-06" }, { "label": "Axios, Generation Lab launches Verb.AI", "href": "https://www.axios.com/2025/05/20/gen-z-data-privacy-generation-lab-verb-ai", "date": "2025-05-20" } ] } ] }