Sell company data to AI companies

Why AI companies want to buy company data

Large language models have already been trained on most of the open web. What is missing is data that is not public: how real companies work, reason and make decisions, and content that is cleared for use. That is why AI companies are increasingly looking for businesses that want to sell data to AI by licensing data they already hold.

Two trends are driving demand:

• AI agents. AI labs are increasingly buying workplace data such as Slack, email and support tickets to train agents that can carry out real work. In the US, shut-down startups have sold their Slack and email archives to AI companies, as reported by Fast Company, Gizmodo and Forbes, among others.
• Gaps in European languages. Smaller European languages make up a tiny share of global training data. Nordic projects such as AI Sweden's GPT-SW3 were built on a dedicated Nordic corpus, OpenEuroLLM has had to generate synthetic data to fill gaps, and the National Library of Norway found that licensed books and newspapers made Norwegian models better.

For you, this means data that is currently just sitting in storage can become a new revenue stream – data monetization without building an AI product of your own. Nordic and European business data, in local languages, is exactly what is hard to find.

Which data can you earn revenue from?

We broker two main types of data.

Operational data

What your day-to-day business produces:

• email archives
• Slack and Teams
• documents and intranets
• CRM history
• support tickets, chat logs and call recordings
• exports from business and accounting systems

Operational data is often valuable because it shows complete sequences – a question, the reasoning, a decision and an outcome. It also almost always contains personal data, which calls for careful preparation.

Content and media archives

Published text, books and manuals, images, video, audio and podcasts, source code and structured datasets. Here the main question is copyright, which is often easier to sort out. Read more about text, image and audio data.

License your data – don't sell it off

When we talk about selling data to AI, we in practice mean licensing data. It is an important difference:

• You keep ownership. The buyer gets a right to use the data, not the data itself forever.
• The license is limited. By default it is non-exclusive and limited in scope, time and use.
• You approve every deal. Nothing is sent without your explicit yes.
• Confidentiality first. Nothing is shared until a non-disclosure agreement and an accepted offer are in place.

Because the license is non-exclusive, the same data can in some cases be licensed to more than one buyer.

Copyright and opt-out

The EU's Digital Single Market (DSM) Directive allows text and data mining of lawfully accessible content unless the rights holder has reserved those rights in an appropriate way – for online content, in a machine-readable form. For content owners, the combination is clear: by reserving your rights against free mining and offering a license at the same time, you move from free use to paid access.

The EU AI Act raises the bar on the buyer side. Under Article 53, providers of general-purpose AI models must respect such reservations and publish a summary of the content used for training. Our documentation is designed to help buyers meet those requirements, which makes your data easier to buy.

Personal data and GDPR

Data that contains personal data takes more work. Data protection authorities stress that reusing data for AI training must be compatible with the original purpose, and the European Data Protection Board has made clear that legitimate interest is not an automatic basis for AI training. That is why we assess rights and personal data first, prioritize data without personal data and plan de-identification before delivery. Health data is outside what we handle.

There is more on this on our page about GDPR and anonymization. What we write is general information, not legal advice – in many cases you will also need your own legal review.

How it works

1. You describe your data in our form. It takes about two minutes and you don't upload any files.
2. We assess rights, personal data and demand, and get back to you within 48 hours on business days.
3. You approve a concrete offer and agreement – or say no.
4. Delivery and payment. The data is de-identified and delivered, and then payment is made.

It costs nothing to have your data assessed. We get paid through a share of the revenue, and only when a deal actually happens.