About the paper
One company, followed every day
OpenAI is written about in hundreds of places each day, and it publishes a good deal about itself as well. OpenAI Daily News collects that reporting, files it by subject and tells you who published it, so that a reader can follow the company without reading everything.
What we are, and are not
The paper is a digest. Most of what appears here was reported by someone else, and each report carries the name of the outlet that published it and a link to the original. We show a headline and a brief excerpt. We do not reproduce anyone's article.
We have no connection with OpenAI. The company does not own, fund, review or endorse this paper, and its trademarks appear only because they are the subject of the news.
Where the reports come from
Every ten minutes we read a fixed list of feeds. From OpenAI itself: its news feed, its research index, its channel on YouTube, its status page, and the release notes for Codex and for the official developer libraries. From the press: about fifty titles, among them The New York Times, The Guardian, BBC News, Bloomberg, CNBC, Axios, The Verge, Ars Technica, TechCrunch, The Register and The Decoder. We also run news searches, which bring in reports from outlets we do not read directly.
The whole list, with the time each source last published, is on the sources page. It is drawn from the latest reading and cannot be edited by hand.
How a report is chosen
Most of those outlets cover the whole of technology or business. A report is kept only if its headline names OpenAI, its chief executive, or one of its products: ChatGPT, a GPT model, Sora or Codex. Shopping offers, share tips, lists of prompts and paid posts are discarded, as are press-release wires.
Each report is then placed in one section according to the words in its headline: ChatGPT, Models & Research, API & Codex, Business, Policy & Courts or Outages. The sorting is done by rule, not by an editor, and it is sometimes wrong. Tell us when it is.
When several outlets carry one story
The same event is often reported by a dozen newsrooms. We show it once, under the fullest version we have, and name the other outlets beneath it. The front page list headed “Most widely reported” is ordered by that count. A story that many independent newsrooms have carried is more likely to be sound than one that appears in a single place, though it is no guarantee.
Posts from OpenAI
Announcements, status notices and release notes that come from OpenAI are marked “Primary source” and link to OpenAI's own pages. They are the company's account of itself and should be read as such. For prices, limits and terms, OpenAI's own pages are the authority, not this paper.
Pieces by readers
The Contributors section carries original essays and notebook entries from registered readers. An editor reads each one before it is published, and the writer's name is on it. These are the only pages here whose text is our own.
Who publishes it
OpenAI Daily News is owned and published by Abraim Sina, who answers for what appears in it. The paper is independent and privately published. It is paid for by the advertisement described on our advertising page, which is always marked as advertising. An advertiser buys space and nothing else: no coverage, and no say in what appears.
Corrections
If a report sits in the wrong section, a credit is wrong, or a reader's piece has a fact wrong, write to the desk through the contact page. We correct errors promptly and say that we have done so.