How to sell company documents and shared drive files for AI training

Shared drives hold the written version of how a company works: SOPs, specifications, proposals, financial models, postmortems and the drafts behind each of them. AI labs value these records because they show professional work in progress, which public text rarely shows. Handshake AI's data partnerships page gives financial models, process documentation, design documentation and specifications as examples of what it buys.

Who buys company documents for AI training?

Company documents are bought by training-data companies, direct buyers and brokers, often as the core of an operational archive.

Our profiles of Handshake AI and Endless Data explain how far their headline figures can be relied on.

How much are company documents worth?

Company documents have no published price per file or per page. The published payouts cover whole archives in which documents are usually the largest part.

• micro1: $100,000 to $500,000 for a qualified partnership, $500,000 to $3 million for large-scale or ongoing datasets
• Mercor: $150,000 to $300,000 for consulting firms' client deliverables, models and the drafts behind them, as reported by ddiy.co in October 2026
• Handshake AI: a headline of $100,000 to more than $4 million, with two calculator estimates on the same page that it does not explain

Smaller archives sit lower: Fortune reported in September 2026, citing SimpleClosure, that closed startups typically receive $10,000 to $100,000 per company across code and data. Treat headlines as ceilings. Our page on document and form data shows how buyers judge what a document archive can teach a model.

What makes a document archive more valuable?

A document archive is more valuable when it keeps the drafts, comments and decisions behind each final version, in native formats, organized by project and year.

Your company owns the copyright in documents its employees wrote as part of their jobs. The US Copyright Office's Circular 30 says a work prepared by an employee within the scope of employment is a work made for hire, and the employer is the author and owner.

Contractors are the gap. A commissioned work counts as made for hire only if it falls in one of nine categories, such as a compilation or instructional text, and both sides signed a written agreement saying so. Otherwise the contractor owns it unless they assigned it to you, so check every consultant and agency agreement.

We are a broker, not a law firm. Take independent advice on any category you are unsure of, and see our guide to whether you can legally sell company data.

How do you export shared drive files?

You export shared drive files with your platform's admin or compliance tools, so you get a fixed copy rather than giving a buyer access to live folders. Google Vault exports Drive files and lets you choose whether to include access-level information, according to Google's Vault help.

SharePoint, OneDrive and Notion have their own export and compliance tools. Whatever you use, keep the folder path, created and modified dates, file type and version history where it can be exported, and record a hash of each file in your data manifest.

What should you strip from documents before export?

You should strip the hidden layers of each document as well as the visible text, because metadata and comments carry names that a reader never sees.

• File properties such as author, last modified by and company
• Reviewer names on comments and tracked changes, replaced with pseudonyms while keeping the edits
• Hidden sheets, rows and cells in spreadsheets
• Headers, footers and cover pages that name clients
• Embedded images, screenshots and signatures in scanned PDFs, with redaction burned into the image
• Links to internal systems and shared folders

Our de-identification guide covers methods, and our vetting standard requires a re-identification test on a random sample.

If you hold UK or EU data

Under section 11 of the UK's Copyright, Designs and Patents Act 1988, an employer owns work employees create in the course of employment, and contractors own their work unless they assign it. Names, emails and other personal data inside documents are covered by UK GDPR and EU GDPR, so remove them or establish a lawful basis.

How do you start selling company documents?

You start by mapping drives and workspaces by owner, project and year, marking folders that are client, HR, legal or third-party material. That map becomes your exclusion list and the first version of the manifest.

You approve every buyer, every sample and every license term, and you can withdraw at any point before signing. Book a free consultation to find out what your document archive could be worth.