23:26 UTC
About

Why ethics.ai exists

Artificial intelligence is being woven into hiring, medicine, warfare, art and childhood — faster than any previous technology. Whether that goes well is not a technical question. It is an ethical one, being answered every day in papers, parliaments, courtrooms and product launches.

ethics.ai keeps the record. Every day we automatically gather and organize what was published across 492 primary sources — research preprints, regulatory announcements, institute reports and journalism — so that anyone can see, in one place, how humanity is handling AI.

What we are

What we are not

We do not host or reproduce full articles, we take no advertising from AI vendors, and we do not editorialize the feed. Abstracts and snippets remain the property of their original publishers.

How this is actually built

We'd rather tell you plainly than have you guess: the daily feed — news, research, policy, field notes — is compiled by software, not a newsroom. A scheduled job fetches each source, and a keyword-based classifier tags every item by topic, company and region; that classifier is deliberately simple and sometimes wrong or incomplete, which is why some pages are thinner than others. The glossary, timeline, frameworks guide, debates library and people directory are different: those are hand-researched and hand-written, checked against primary sources, and updated on a slower editorial cycle rather than a daily cron.

Pages with very little matched content are marked so search engines don't index them as if they were fully-populated trackers — we'd rather a page be honestly thin than falsely look comprehensive.

Corrections and contact

Automated classification will misfire sometimes — a company page missing an obvious story, a topic tag that doesn't fit. If you spot an error, in either the daily feed or the hand-written reference pages, or you're a publisher who wants a feed added or removed, write to corrections@ethics.ai. Corrections to factual claims in the reference pages are logged and dated.

For machines

Structured access: RSS · JSON API · llms.txt (index) · llms-full.txt (full inlined reference content). Attribution required: "via ethics.ai".