Latest articles

About the archive

This is the complete archive of Data Today, newest first. Every article on the site is listed here, whichever desk produced it: AI and research, data engineering, security, business and markets, or opinion and explainers.

Data Today covers data and AI for people who build and run systems with the technology. Each article starts from a primary source, such as a model card, a research paper, a filing, release notes, or a dataset, and turns the numbers in it into a clear read on what they mean for your codebase, your costs, or your roadmap, with a labelled chart when the data supports one. Articles are drafted by our research pipeline and reviewed, edited, and signed off by the responsible editor, Lars Cornelissen, before they go live. The editorial policy explains that division of labour in detail.

How to use the archive:

• The search box filters the list by title, section, and summary as you type.
• The section pages in the main navigation narrow it to one beat.
• Each desk page under authors and desks lists that desk's work.

We publish at most 25 articles a month. That limit is deliberate: it is the volume one editor can check properly, and drafts beyond it wait in a queue rather than going straight to the site. When two articles cover the same story, the weaker one is folded into the stronger and its old address redirects there.

Older articles carry an "Updated" date where something changed after publication, such as the move to desk bylines in October 2026. The original publication date always stays as it was.

If you find an error in any article, the corrections address is on the contact page, and every substantive fix is logged on the corrections page.