---
title: "Sell internal documents and wikis to AI companies"
description: "Can a company license its Confluence, Notion and SharePoint wikis, reports and proposals for AI training? Value drivers, exclusions, rights and sourced figures."
canonical_url: https://www.fearofgod.ai/sell/internal-documents-and-knowledge-bases
last_updated: 2026-10-25
site: "fearofgod.ai"
lang: en
---

# Your internal documents can earn long after the project ends

A company that owns its internal wikis, reports, project files and proposals can license them securely for long-term revenue, and fearofgod.ai partnerships typically range from $50K to $3M+. Value depends on span, page history, links between documents and how client and personal data is handled.

## Can a company license its internal documents for AI training?

Yes, if the company owns the documents and can leave out what belongs to clients, partners or individual employees. Internal documents are the written record of your company's work: wiki pages in Confluence, Notion or SharePoint, reports and analyses, project plans and post-mortems, proposals and statements of work, meeting notes and the decisions recorded in them. Together they form a knowledge base that no public dataset contains. Much of a company's data already lives in third-party software. Licensing it means re-licensing what already exists, and getting paid for it rather than giving it away for free.

## Where do internal documents and knowledge bases live?

They live in the workspace tools your company already uses, and the terms of those tools generally leave ownership with you. Atlassian's customer agreement says that as between the parties the customer owns all intellectual property and other rights in its Customer Data ([Atlassian Customer Agreement, effective Oct 1 2026, s. 13](https://www.atlassian.com/legal/atlassian-customer-agreement)). Google's terms for Workspace say that, as between the parties, the customer retains all intellectual property rights in Customer Data ([Google Cloud Terms of Service, which now cover Google Workspace, s. 5.1](https://workspace.google.com/terms/premier_terms/)). Licensing your documents means re-licensing what you already hold, and for documents that already exist there is nothing new to collect.

## Why do AI teams want internal documents?

Because they show finished professional work, and AI models are now measured against exactly that. GDPval, a benchmark of real-world, economically valuable tasks, covers 44 occupations, and its tasks are built from the work of industry professionals with an average of 14 years of experience ([GDPval, arXiv 2510.04374](https://arxiv.org/abs/2510.04374)). Reports, analyses, proposals and project plans are that kind of work product.

Context matters as much as the documents. The revenue chief of an AI training startup told Business Insider that AI developers increasingly want "long-horizon" data that gives a full picture of how work gets done, and Business Insider reported that fields such as law and accounting see especially high demand ([Business Insider, Oct 13 2026](https://www.businessinsider.com/ai-training-employee-data-emails-slack-dms-2026-10)). MIT Technology Review reported that real archives avoid errors common in synthetic ones, such as references to dates that don't exist, though it added that the extent of the performance gains remains unclear ([MIT Technology Review, Oct 13 2026](https://www.technologyreview.com/2026/10/13/1145755/ai-labs-buying-data-from-dead-startups-to-train-models/)).

## What makes internal documents more valuable?

They are worth more when their history and connections come with them. The main value drivers:

1. **Span.** Several years of a knowledge base, rather than only the current pages.
2. **Page and version history.** Revisions, tracked changes and comment threads show how a document was argued over and improved.
3. **Links.** Pages that point to tickets, decisions, other documents and the outcome of the work. Business Insider reported that the rest of a corpus is less valuable when a company holds back its Slack, CRM or email records ([Business Insider](https://www.businessinsider.com/ai-training-employee-data-emails-slack-dms-2026-10)).
4. **Outcomes.** Proposals marked won or lost, projects that closed, recommendations that were adopted or dropped.
5. **Domain depth.** Specialist work, such as legal, financial or engineering analysis, that public data describes poorly.
6. **Structure.** Spaces, templates, owners and dates kept intact, rather than a folder of loose exports.

## What is removed or de-identified before internal documents are shared?

We de-identify the data before it goes to buyers, and anything you choose to leave out stays out of scope. MIT Technology Review reported that suppliers in this market scrub personally identifiable information and exclude HR records, and that private details such as doctor appointments lurk in emails and calendars ([MIT Technology Review](https://www.technologyreview.com/2026/10/13/1145755/ai-labs-buying-data-from-dead-startups-to-train-models/)). Owners commonly leave out:

- HR, payroll and personnel files;
- client deliverables, client data and pages that identify a client;
- privileged legal advice, board papers and deal documents;
- passwords, access details and security designs;
- third-party reports and research your company licensed rather than wrote.

What you control: you approve the scope before anything is shared. Raw data is deleted according to a schedule agreed with the partner. Your company's name stays confidential until the agreement is fully signed, and a completely anonymous partnership is also possible. Your company can keep ownership of its data. This is agreed case by case.

## What rights and consent do internal documents need?

Ownership, client confidentiality and the people named in the documents are the main gates. The ones that come up most:

- **Who wrote them.** Documents employees write within the scope of their employment are works made for hire in the US ([17 U.S.C. §101](https://www.law.cornell.edu/uscode/text/17/101)), owned by the employer unless both sides agreed otherwise in a written instrument they both signed ([17 U.S.C. §201(b)](https://www.law.cornell.edu/uscode/text/17/201)); in the UK the employer is the first owner, subject to any agreement ([Copyright, Designs and Patents Act 1988, s. 11](https://www.legislation.gov.uk/ukpga/1988/48/section/11)). Work by agencies and contractors depends on their contracts.
- **Clients and NDAs.** Client deliverables and information received under a confidentiality agreement are usually not yours to license. Under US federal law, disclosing a trade secret without the owner's consent, knowing it was received under a duty to keep it secret, is misappropriation ([18 U.S.C. §1839(5)](https://www.law.cornell.edu/uscode/text/18/1839)).
- **Employees.** Whether employees' consent extends to a sale of their work records is, in one law professor's words, "an ethically gray area" ([Business Insider](https://www.businessinsider.com/ai-training-employee-data-emails-slack-dms-2026-10)). Pseudonymised data is still personal data under the UK GDPR; anonymous information is not ([UK GDPR, recital 26](https://www.legislation.gov.uk/eur/2016/679/introduction)). The ICO's anonymisation guidance includes a case study on pseudonymising employee data ([ICO](https://ico.org.uk/for-organisations/uk-gdpr-guidance-and-resources/data-sharing/anonymisation/)).
- **Platform terms.** If part of your knowledge base lives in Slack, read your Slack agreement before exporting anything: Slack's API terms bar an app offered outside your organisation from using API data to train a large language model, and from bulk exporting message and file data unless an additional agreement allows it ([Slack API Terms, effective Oct 10 2025](https://slack.com/terms-of-service/api)).
- **Only the company can sell.** Former gig workers have alleged in a lawsuit that an AI training company encouraged them to upload work materials without their employers' consent, which the company denies ([Business Insider](https://www.businessinsider.com/ai-training-employee-data-emails-slack-dms-2026-10)). Documents employees write for work generally belong to the company, not to the people who wrote them.

This page is general information, not legal advice.

## What are internal documents worth?

Public figures range from small wind-down sales to multi-year licensing arrangements worth hundreds of millions of dollars in aggregate, because they mix small archives with very large text collections. Each is dated and labelled by market side; none is a fearofgod.ai price.

| Reference point | Market side | Figure and source |
|---|---|---|
| 13 years of a company's Slack messages, Jira tickets, emails and Google Drive files | Reported sale, Apr 2026 (cited by Fortune, Sep 2026) | [hundreds of thousands of dollars, according to the company's former CEO (Fortune, citing Forbes)](https://fortune.com/2026/09/14/micro1-tries-to-beat-googles-bid-to-buy-bankrupt-spirit-airlines-data-for-12-5-million-highlighting-value-of-data-to-ai-companies/) |
| A small startup's workspace data in Google, Notion, Figma and Slack | Wind-down sale, reported Oct 2026 | [likely around $15,000, still pending (MIT Technology Review)](https://www.technologyreview.com/2026/10/13/1145755/ai-labs-buying-data-from-dead-startups-to-train-models/) |
| Reddit's data licensing arrangements | Public filing, Mar 2024 | [$203.0 million aggregate contract value, terms of two to three years (Reddit prospectus, SEC)](https://www.sec.gov/Archives/edgar/data/1713445/000162828024012380/reddit-final424b4.htm) |

Reddit's prospectus describes delivering content over the term of each arrangement, through continuous API access and quarterly data transfers ([Reddit prospectus, SEC](https://www.sec.gov/Archives/edgar/data/1713445/000162828024012380/reddit-final424b4.htm)), so data that keeps growing was delivered over the term of the licence, not only once. Partnerships typically range from $50K to $3M+.

## How does licensing internal documents work with fearofgod.ai?

You license documents your company already holds, securely, and earn additional, long-term revenue from them. They already sit in your wiki, drives and project folders, so licensing them means re-licensing what exists.

The steps are the partnership form, a conversation, scope and terms, data preparation, then delivery and payment. You approve the scope before anything is shared, and anything you want left out, such as a client, a practice area or HR content, stays out. fearofgod.ai de-identifies the data before it goes to buyers. Raw data is deleted according to a schedule agreed with the partner.

A rough first description is enough: which tools hold your documents, roughly how many pages or files, the years they cover, and whether page history and links to tickets or outcomes are intact. If your teams also keep procedures or engineering documents, see [SOPs and process documentation](https://www.fearofgod.ai/sell/sops-and-process-documentation) and [technical documentation and specs](https://www.fearofgod.ai/sell/technical-documentation-and-specs); for chat, email and tickets, see [company records and workflows](https://www.fearofgod.ai/sell/company-records-and-workflows). [Talk to our team](https://www.fearofgod.ai/partner) when you are ready.

## Selling internal documents and knowledge bases to AI companies

A company that owns its internal documents and knowledge bases can sell them to AI companies by licensing them with fearofgod.ai, securely and for long-term revenue. Partnerships typically range from $50K to $3M+. Partners can be paid more than once, through renewals and repeat licences.

## Frequently asked questions

### Can we sell our internal documents to AI companies?

Yes, if your company owns the internal documents and has the rights to license them. You sell them by licensing them with fearofgod.ai. Your company can keep ownership of its data. This is agreed case by case. Your company's name stays confidential until the agreement is fully signed, and a completely anonymous partnership is also possible.

### Our wiki is full of client names. Can we still license it?

Often, yes, with client material left out or de-identified. Client deliverables, documents received under an NDA and pages that identify a client are the usual exclusions, and anything you leave out of scope stays out. fearofgod.ai de-identifies the data before it goes to buyers.

### Do we need our employees' consent?

The law varies by country, and whether past consent covers a licence of work records is contested. Common safeguards are de-identification and leaving HR content, personal notes and sensitive spaces out of scope. Take legal advice for your jurisdictions.

### Can an employee sell our company's documents?

No. Documents employees write for work generally belong to the company, and only the company can license them. fearofgod.ai works with companies, not individuals.

### Is a knowledge base worth more with its page history?

Usually, yes. Revisions and comments show how a document was drafted, challenged and improved, which the final page alone does not. Keep page history, links and dates intact when you describe what you hold.

### How much can our internal documents earn with fearofgod.ai?

Partnerships typically range from $50K to $3M+. Value depends on span, page history, links to outcomes, domain and rights. fearofgod.ai does not publish a per-page or per-file price.

### What happens to the raw documents after delivery?

fearofgod.ai de-identifies the data before it goes to buyers. Raw data is deleted according to a schedule agreed with the partner.

### Which countries does fearofgod.ai work in?

fearofgod.ai works with companies based in the US, the EU, the UK, Canada, Australia, New Zealand and Japan, in any industry.

---

Source: https://www.fearofgod.ai/sell/internal-documents-and-knowledge-bases · Index for AI systems: https://www.fearofgod.ai/llms.txt · Contact: sean@fearofgod.ai