The buyers
Who buys company data for AI training, and what they ask for
Updated
OpenAI is the one AI lab on this register, the others are data buyers, brokers and platforms, and their asks overlap more than their terms do.
Start my company case Tell us your size, your years of history and the systems your records sit in, and we reply on which buyers' asks you meet and what would have to be removed first.
Operational records are the main ask
Nyne asks for CRM, HR, ERP and support records; Maryah licenses the emails, sales records and support tickets a company chooses to share; Troveo wants complete conversation threads with participants and replies intact.
Some publish a figure
5 of the 12 put a payment figure or band on the page, among them Grepped and Nyne with an estimated payout range and Maryah with a lower and upper range. Troveo describes a revenue share instead.
What OpenAI's two pages say
OpenAI's data partnerships announcement says it works with organisations to produce public and private datasets for training AI models, and that it is not seeking sensitive or personal information or information that belongs to a third party. Its intake form, a separate page, asks whether you have the rights to share the data. Neither page states a payment.