How do you license your company's data to AI buyers?

Updated Oct 19, 2026 · Sources linked below · By DataBrokering.com

Short answerYou license company data to an AI buyer by confirming you have the right to share it, deciding what to exclude, getting introduced to a buyer that fits your size and data, signing an NDA, letting the buyer review samples and make its own offer, then signing a license that controls de-identification and use. Buyers set the price; nobody can honestly quote one up front.

Question 1 of 9: Is your company active or shutting down?

Active →Shutting down / shut down →Free 2-minute fit check. We never ask for files, logins or data.

What AI buyers are looking for

Buyers want records of how work actually gets done. micro1's program page lists SOPs, knowledge bases, internal documentation, CRM data, project histories and QA processes, and its FAQ says the most valuable datasets “capture real operational expertise, decision-making processes, and repeatable workflows.”¹ MIT Technology Review reports that wind-down sales cover code plus “workspace data—emails, chats, documents.”³

Demand is growing. Business Insider reports Gartner's estimate that annual spending on data for AI will grow from under $1 billion to almost $6.5 billion in 2027, and that law and accounting “see especially high demand.”²

Who qualifies

Active companies. micro1 says it looks for “operationally mature companies with 30+ employees,” prioritizes US companies “followed by other Western markets,” and is primarily looking for companies whose documentation is in English.¹ Other buyers set different size and industry rules, and some accept smaller professional-services firms. See the qualification checklist.
Companies shutting down. SimpleClosure's Asset Hub has handled over 150 deals, and 47% of SimpleClosure deals now include code or data sales, per MIT Technology Review.³ See selling data when you shut down.

The steps, in order

Inventory what you hold. List each dataset by system, record count and start date, and mark each as included or excluded.⁴
Confirm you can sell it. Review customer, partner, vendor and data license agreements, plus the platform terms of the systems you'd export from.⁴
Define exclusions. Frankfurt Kurnit advises excluding high-risk data entirely rather than relying on de-identification: privileged material, HR and personnel files, payroll and tax records, health information, union-related communications, and customer or partner confidential information.⁴
Get introduced to one buyer that fits. Buyers differ on size, industry, geography and whether they take shut-down companies.
NDA, then evaluation. micro1 describes its process as an assessment, an evaluation of fit, then a discovery call about your data and privacy requirements before any agreement.¹ Don't share sample data before your own review: “a sample itself could be a disclosure.”⁴
Offer and license. The buyer sets its offer. The agreement should restrict how the buyer can use the data, allocate privacy risk, and say who de-identifies the data and how.⁴

“Selling company data for AI training can be done, but most companies cannot sell everything a buyer requests, and the work required is usually more than the offer suggests.”Daniel Goldberg, Partner, Frankfurt Kurnit Klein & Selz⁴

The main risks

Rights. Holding data doesn't mean you have the right to sell it. Email and chat contain other companies' confidential information.⁴
Privacy promises and law. You're bound by the privacy statements in effect when the data was collected, and under the CCPA a paid transfer of workplace personal information is a sale unless the data meets the statute's de-identification standard.⁴
Permanence. Once data is used to train a model, “it cannot practically be pulled back out.”⁴
Pressure tactics. Business Insider reports owners receiving offers that expire in 48 hours. See red flags in AI data offers.²

FAQ

Can any company license its data to AI buyers?

Not every company. Each buyer sets its own rules on size, industry, location and language. micro1, for example, publishes a 30+ employee minimum and prioritizes US companies with English documentation. Companies that are shutting down can also sell code and workspace data through wind-down services.

How much is company data worth to AI buyers?

There's no reliable public price. Each buyer reviews the data and makes its own offer based on factors like volume, quality and uniqueness. MIT Technology Review reports that buyers quote wildly different numbers for similar data, so be wary of anyone who quotes a figure before seeing samples.

Do I have to hand over my data to DataBrokering.com?

No. DataBrokering.com makes introductions only. We never ask for files, logins or exports. Any data review happens directly between your company and the buyer.

What data should I exclude?

Legal guidance from Frankfurt Kurnit suggests excluding privileged material, HR and personnel files, payroll and tax records, health and benefits information, union-related communications, customer and partner confidential information, licensed data and data from regulated systems.

How we’re paid

The fit check is free for companies, and companies never pay us. Some buyers pay DataBrokering.com a referral fee when a company we introduce signs with them. We tell you before any introduction whether a fee may be paid. Buyers set their own offers, payment terms and timelines. We don’t quote, estimate or promise amounts.

See which buyer fits your company.
9 quick questions, then one introduction. Free for companies.

Check if my records qualify →

Does my company qualify to license its data to AI? A 9-point checklist Got a 48-hour offer for your company's data? 8 red flags to check first Can a law firm license its data to AI buyers?Can an accounting or CPA firm license its data to AI buyers?Shutting down your startup? You may be able to sell its Slack, email and code data

Sources

micro1, “Monetize your company's operational data” (program page and FAQ). https://www.micro1.ai/data-partnerships
Business Insider, “The AI Data Grab” series: Rollet, Carter & Council (Oct 13, 2026). https://www.businessinsider.com/ai-training-employee-data-emails-slack-dms-2026-10
Matthew King, MIT Technology Review (Oct 13, 2026). https://www.technologyreview.com/2026/10/13/1145755/ai-labs-buying-data-from-dead-startups-to-train-models/
Daniel Goldberg, Frankfurt Kurnit Klein & Selz, “Selling Company Data for AI Training: What to Know Before You Say Yes” (Oct 1, 2026). https://technologylaw.fkks.com/post/102o455/selling-company-data-for-ai-training-what-to-know-before-you-say-yes