Scale AI: Data Partnerships
Scale AI Data Partnerships license de-identified records from a company's business systems. 40+ employees preferred, illustrative value of $10K to $1M+.
Published October 19, 2026. Last checked October 20, 2026
What Scale AI takes
- •Structured, de-identified records from data warehouses (BigQuery, Snowflake, Databricks)
- •Chat, email and calendar records (Slack, Teams, Google Chat, Gmail, Outlook, Google Calendar)
- •Documents (Word, Docs, Slides, Excel, Sheets, Drive, SharePoint)
- •Work tools (Jira, Confluence, Asana, Notion, Linear)
- •Business platforms (Salesforce, HubSpot, Workday, ServiceNow, SAP, Zendesk)
What Scale AI asks for
- •40+ employees preferred, assessed case by case
- •Established data systems and real operational workflows
- •Domain depth
- •Someone inside the company who owns the partnership
What Scale AI pays
- •$10K to $1M+ per partnership, scaling with cadence. Scale's own illustration, not a quote
From Scale AI's own pages, checked October 20, 2026.
A few questions, about two minutes. You see which buyers fit, Scale AI included.
Details on who qualifies
“40+ employees is our preferred starting point”, Scale says, and it assesses every partnership case by case. The workflows have to be real: demos and synthetic data don’t count.
The industries on Scale’s page are software and technology (including telecoms, managed IT, marketing agencies and direct-to-consumer e-commerce); financial, legal and insurance; healthcare and science; industrial, energy and logistics; and public, services and consumer (retail, staffing, hospitality).
Details on pay
A partnership is continuous, and Scale says it can begin with a very limited scope.
What Scale does with your records
Scale anonymizes records inside your environment before anything is extracted, and you approve each package before it leaves. Scale says data is encrypted in transit and at rest, logged in an audit trail, and available with on-premises and VPC options. Employee and customer identities, raw systems, credentials and source access stay with the company.
How it works
- Apply. Start a partnership or book a scoping call, both from Scale’s page.
- Scoping. You and Scale agree which systems are in scope.
- Agreement. Scale signs a partnership agreement with the companies it takes on.
- Packages. Records are de-identified inside your environment, and you approve each package before it is extracted.
- Later packages. The partnership continues, and later packages can widen the scope.
Before you apply
- Name the person who will own the partnership, and confirm who can sign a continuing agreement.
- Keep the systems running. De-identification happens inside your environment, so Scale needs live access to Slack, the warehouse, the CRM and the rest. If the company is closing, raise this on the scoping call before you cancel anything; see selling company data after a shutdown.
- Know which systems hold customer personal data. You approve each package, so someone on your side has to recognize what is in it.
- Read customer contracts and data processing agreements for clauses that restrict sharing records that mention customers, even de-identified ones.
Sources
- Scale AI, Data Partnerships checked October 20, 2026
Related guides
- Acqui-hired? What happens to the old company's records After an acqui-hire the old company often keeps its Slack, tickets, docs and some code. Who owns them now, who can sign, and which buyers license them.
- Can a company in administration or bankruptcy sell its data? Usually yes, if the administrator, trustee or assignee signs. Who has authority, why customer personal data stays out, and which buyers license the records.
- Sell company data after shutdown: what to keep, who buys A closed company can still license its Slack, email, tickets and docs to AI buyers. Which buyers take closed companies, and what to export before you cancel.