Research
Research papers
Working papers by AIData.Broker on the market for AI training data. Every figure comes from the open datasets, where each row names its source.
18 Oct 2026 · Yuma Heymans
AI data licensing deals, July 2023 to October 2026: a public dataset and what it shows
Licences of content to developers of artificial intelligence (AI) are negotiated privately and recorded in no registry; what is known comes from announcements, securities filings and press reports. This paper describes an open dataset that compiles those disclosures, with a source link, date and verbatim quote for every row, and reports what it shows....
18 Oct 2026 · Yuma Heymans
Since 2026 AI data vendors, marketplaces and intermediaries have published programmes through which an ordinary company can sell or license its internal records for training AI models. This paper compares the published terms of 37 organisations with a public way in, as compiled in an open dataset reviewed on 16 October 2026, together with a register of...
18 Oct 2026 · Yuma Heymans
When a software company closes, its code, tickets, documents and chat outlive it, and since 2026 AI data vendors and wind-down services have offered to buy them. This paper describes a public register of technology companies that are closing, sold or have gone dark, compiled each morning from public signals, and sets it beside the public record of what...
18 Oct 2026 · Yuma Heymans
De-identifying company workspace exports in the data holder's browser: design and benchmark results
Company records such as chat workspaces, mailboxes, documents and tables are proposed as training data for language models, and they carry identifiers of employees, customers and other organisations. This report describes a rule-based de-identification engine that runs entirely in the data holder's browser, so that raw records, the people learned from...
18 Oct 2026 · Yuma Heymans
How AI developers acquire training data: six channels and the shift away from the open web
Developers of large AI models obtain training data through six channels: collection from the open web without a licence, licences from publishers and platforms, commissioned human and expert work, the purchase or licence of organisations' internal records, purpose-built environments and synthetic data, and data from their own users and devices. This...
18 Oct 2026 · Yuma Heymans
Developers of AI models now license the working records of ordinary companies: support tickets, email, chat, CRM activity, documents and code. Whether a company may license such records is not one question but several, and buyers' counsel ask them in a recognisable order. This article compiles what public sources show on those questions for the United...
18 Oct 2026 · Yuma Heymans
What AI training data sells for: a sourced index of public prices, 2023 to 2026
Training data for machine learning is bought and sold without an exchange or a posted price list. This paper assembles the public record of its prices: 1,022 observations for nine kinds of data, read from 735 source documents between 11 and 17 October 2026, 1,009 of them dated 2023 or later. Each observation keeps the figure as published, the side of the...
18 Oct 2026 · Yuma Heymans
Website operators who object to the use of their pages for training AI models usually say so in robots.txt, the plain-text file that tells automated clients what they may fetch. This paper reports a census of those files. A reader takes the Tranco ranking of popular sites, keeps registrable domains with eleven address endings within the top 300,000...
18 Oct 2026 · Yuma Heymans
Firms that sell training data, expert labour, reinforcement-learning (RL) environments and evaluations to AI developers are mostly private and known largely through their own websites. This paper reports a census of 150 such vendors, read on their sites on 15 and 16 October 2026 with every claim tied to a quoted sentence, and links it to three companion...