Buyers of scientific data
Licensing research data to AI: who takes it, and how
Updated
The organisations that talk about scientific data for AI fall into four groups, and only one takes records from the companies and labs that hold them. This is what each says on its own page.
Those that take data from labs and companies
Nyne says it partners with companies to license their data, and Troveo that it anonymises a company's data to a documented standard and licenses it on terms the company approves. Neither singles out lab records, so a lab comes to them as a company with operating data.
Source: Scientific data buyers
Those that license their own corpus, or build one
Springer Nature's ARC3 licensing, Wiley's AI data products and Redpine's licensed peer-reviewed research are published literature, not a lab's notebooks. Lila Sciences says its autonomous labs design and run experiments and learn from the new data, and K-Dense has partnered with Litmus Science for on-demand lab work: both make data rather than buy it. Polaris is where drug-discovery datasets and benchmarks are shared.
Source: Scientific data buyers
AWS Data Exchange and Opendatabay are venues where a holder lists data for buyers to find. If you are selling the whole lab, see selling a research lab.